UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Complex formation between heterocyclic compounds and polynitro benzene derivatives McEwen, Kathleen Lenore 1953

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

Item Metadata


831-UBC_1953_A6_7 M3 C6.pdf [ 3.33MB ]
JSON: 831-1.0062294.json
JSON-LD: 831-1.0062294-ld.json
RDF/XML (Pretty): 831-1.0062294-rdf.xml
RDF/JSON: 831-1.0062294-rdf.json
Turtle: 831-1.0062294-turtle.txt
N-Triples: 831-1.0062294-rdf-ntriples.txt
Original Record: 831-1.0062294-source.json
Full Text

Full Text

COMPLEX FORMATION BETWEEN HETEROCYCLIC COMPOUNDS AND POLYNITRO BENZENE DERIVATIVES by KATHLEEN LENORE McEWEN A Thesis Submitted in Partial Fulfilment of the Requirements for the Degree of MASTER OF SCIENCE in the Department of CHEMISTRY We accept this thesis as conforming to the standard required from candidates for the Degree of MASTER OF SCIENCE -University of British Columbia April 1953 Abstract Complexes formed between heterocyclic compounds and polynitro benzene derivatives have been examined spec-troscopically. It has been found that many complexes, when irradiated in the region of their charge-transfer absorption emit a short-lived radiation in the region of the hetero-cyclic triplet. There is some evidence for a radiationless transition from an excited, bonding charge-transfer state to the excited triplet level of the heterocyclic component. ACKNOWLEDGMENT I wish to express my sincere appreciation to Dr. C. Reid for suggesting and advising me in this research. I should also like to acknowledge my indebtedness to the National Research Council of Canada for a studentship. TABLE OF CONTENTS INTRODUCTION -page. The Importance of Molecular Complexes - - - - 1 Simple Types of Molecular Complex - - - - - - 2 The Type of Bonding Responsible - - — - - - 3 The absorption Spectra of Complexes - - - - - $ Molecular Orbital Theory for Complexes - ,9 Phenomena not Immediately Explainable by Charge Transfer. - - - - - - 12 Phenomena Explained by Complex Formation - — 1 3 Reason for present Investigation- - - - - - - 15 A Review of Electronic Properties of Heterocyclics- - 15 Transitions Found in Heterocyclics - - - - - 17 EXPERIMENTAL METHODS AND RESULTS -Apparatus - - - - - - - - - - - - - - - - - - 2 1 Rrocedure - - - - - - - - - - - - - - _ - _ _ 2 1 Results - - - - - - - - 24 DISCUSSION 39a BIBLIOGRAPHY - - 47 APPENDIX 51 COMPLEX FORMATION BETWEEN HETEROCYCLIC COMPOUNDS AND POLYNITRO BENZENE DERIVATIVES Introduction THE IMPORTANCE OF MOLECULAR COMPLEXES The group of compounds known as molecular compounds or molecular complexes has become of great interest during recent years. From a theoretical standpoint alone, i t is extremely important to understand the type of bonding in-volved in complex formation and the various energy relation-ships possible between the components of a complex. However, such knowledge would also add considerably to our understand-ing of many vitally important but complicated reactions in which complex formation is involved at some stage. For instance, i t has been suggested that photosyn-thesis may depend upon the formation of a loose complex between a chlorophyll molecule and a carbon dioxide molecule, in contact with a suitable enzyme. This might solve the problem of how the reaction forming glucose from carbon dioxide and water, which requires 112,000 calories per carbon dioxide molecule, can take place in a system absorb-ing only visible light. There may be a mechanism whereby a chlorophyll molecule could absorb three quanta of red light, which together would correspond to about 112,000 calories, and transfer them to one place where they would be available to the complexed carbon dioxide. -2-Another biological problem, a partial solution to which may l i e in increased knowledge of complex formation, is that of carcinogenesis. It has been found that aromatic carcinogens combine with desoxyribose nucleic acid in the cells, and i t is possible that they may thus change the properties of the acid enough to cause chromosome aberrations© Furthermore, i t is thought that certain compounds, such as caffein may be anitcarcinogenic, since they compete to 17 complex the aromatic carcinogens. SIMPLE TYPES OF MOLECULAR COMPLEX At the present stage of knowledge, perhaps the best way to gain an understanding of the very complicated systems discussed in the foregoing section, will be to turn to com-plexes between far simpler molecules and to review what is known of them. Probably the best known of the organic molecular complexes is the familiar picrate or polynitrobenzene derivative formed by aromatic hydrocarbons»^ The instant deepening of colour observed when a solution containing a hydrocarbon is added to one containing a polynitro compound has been a source of interest for many years. Such complexes are also fonmed between certain quinones and aromatic hydro-carbons .53 In a l l cases the stability of the complex increases with increasing deactivation of the nitro-compound ring, or of the quinone; accomplished, for example, by the addition -3~ of more nitro or carboxy substituents to the former, and halogen substituents to the latter. In respect to the hydro-carbon component, however, stability is increased by activat-ing the aromatic ring system; for instance, by methylation, or by extending the aromatic system over more rings. Another interesting series of molecular complexes which have been examined quite thoroughly during the past few years, are those formed by halogen molecules and aroma-tic hydrocarbons; usually in a ratio of 1 : 1, although there is some evidence for 1 : 2 complexes.^ The f i r s t evidence that complex formation of this type occurs was from low temperature phase diagrams, and from viscosity measurements.^ More recently a number of investigators have examined the absorption peaks of these complexes and have found that, in analogy with the strong colour of the picrates, complex form-ation between a halogen molecule and a molecule of an aromatic hydrocarbon produces an absorption maximum that belongs to neither component. This new absorption band will be discussed in detail later. THE TIEE OF BONDING RESPONSIBLE FOR COMPLEX FORMATION Naturally much curiosity was aroused as to the type of bonding between these pairs of stable molecules. One of the earliest postulates in this connection was that of Weiss, that complexes of quinones or polynitr© compounds with aroma-tic hydrocarbons are simple ionic compounds, in which the hydrocarbon gives an electron to the more electropositive -4-nitro compound or quinone. Weiss presented dipole moment data which indicated "some such bonding mechanism* Soon after this, Woodward, coming closer to the present idea of bonding in molecular complexes, suggested that an intermolecular semipolar bond might explain complex formation; that i s , an arrangement such as (0 (NC^)^/ (Arf*.^9 Dewar has pointed out that London dispersion forces may be used to account for complex formation between aromatic ring systems.^ That is to say, the periodically varying electron densities on the two rings, the time averages of which are represented by the wave functions \t and will synchronize with one another in such a way as to produce a net attraction. This attraction will be strong i f i t results in compensating for a net electropositivity of one ring system. For instance, consider a pair of compounds such as anthracene and s-trinitrobenzene. The ff electron system of the s- t r i -nitrobenzene ring is to some extent electropositive, because of the resonance effect of the nitro groups. The anthracene, on the other hand, has a readily polarizable ^electron system, and may be: Visualized as ready to neutralize the positive charge on the s-trinitrobenzene ring. In a sense the latter may be called an electron acceptor, and the hydrocarbon a donor.^ Very recently, Mulliken put this idea of a partial transfer of an electron into a more exact mathematical form than is implicit in the descriptions of bonding given above. The details of this theoretical approach will be considered 1 in a subsequent section; but, in brief, the theory involves assuming a small amount of ^ (NG^)* - (Ar)~ character in the wave function, in which electron transfer has occurred and the orbitals now containing only one electron overlap to form a covalent bond* This modified Lewis acid-base theory has been borne out by a wealth of experimental evidence, mainly concerning halogen-benzene type complexes* For instance, toluene, m-sylene, and naphthalene form stronger complexes than does 2 c i i benzene, * * and i t is known that the electron densities of the ring systems of the two former are higher than that of the benzene ring; while naphthalene, being much more polariz-able than benzene,will also act more readily as an electron donor. Turning now to the "acid" or acceptor molecule, we find that the strength of the complex increases in order Clg ^ SO2 Br 2 ^ I2 * ICI.3 This is what one would expect, at least for the series Gl 2, Br 2, I2; for clearly the electron affinity, or tendency to become X2" will increase in going from GI2 to I2. The fact that IC1 does not act like Br 2 in regard to strength of complex formation, as would be expected, must be explained by other factors* ' It will be of use in visualizing the form of more com-plicated complexes, to consider the probable electronic struct-ure of the benzene-iodine complex, and to see what factors are important in such a consideration.The most probable form of the complex is that with the iodine molecule resting on the benzene molecule, its axis parallel to one of the twofold axes -6-of the benzene ring, and its centre on the six-fold axis of the benzene. In this form the eomplex will have symmetry ^2v ^ e m u s t n o w examine the molecular orbitals of benzene and of iodine (or I2") under a perturbing field of symmetry ,C2v» There are three ^"molecular orbitals in the ground state of benzene to be considered; while the 5p orbitals of the iodine atom combine to form six molecular orbitals of the form shown in Fig, 1 Fig, 1. Schematic diagrams for the case of a perturb-ing field of symmetry with xy and yz as symmetry flanes. a) Benzene - the relative sizes of the circles indicate the relative charge densities on each atom for that molecular orbital. Shaded circles indicate a positive wave function on the positive-z side of the benzene plane. (b) Iodine - shaded portions indicate positive regions of the wave function, unshaded portions, negative. The y axis is the intermolecular axis. The inner nodes of the atomic orbitals are omitted. -7-The a^, b^, etc, refer to the symmetry species classifications of the orbitals. The ^)t\> ^ refer to the 5p atomic orbitals parallel to the x, y, and z axes respect-ively. In the ground state of iodine there are two elect-rons in each orbital except the last, which is strongly antibonding. However, the ground state of I 2 " will have one electron in the b£ & u orbital, giving a 2 B 2 state for l 2 ~ • Now since the ground state of the complex is required to have the symmetry A^^ , the electron donated by the benzene will have to come from the orbital belonging to the symmetry species b 2 , so that the (benzene)* will also be in its 2B 2 state. It will be seen that this arrangement will give excellent overlap of the two bonding orbitals of I 2 ~ and (benzene)*. The symmetry requirements mentioned above will also hold for nitro-compound complexes as long as the com-plex has some overall symmetry. If i t has not, then the same approximate symmetries must exist for the two com-plexing molecules in the region of their interaction at least. However, i t is clear that such considerations will become quite complicated even in reasonably simple poly-nitrobenzene-hydrocarbon complexes. THE ABSORPTION SPECTRA OF MOLECULAR COMPLEXES As was mentioned earlier, both picrate-type complexes and halogen complexes show absorption bands which are peculiar to neither of the original components. The picrates have a long, continuous absorption from the upper wavelength limit of the hydrocarbon up to 4000 or 5000 A, while the halogen complexes have an intense absorption band at about 3000 A. In changing from benzene to mesity-lene the maximum of this absorption band in SO2 complexes varies from 2800 to 305© A, increasing methylation of the benzene type component having the effect of shifting the band toward the visible. On the other hand, complexes with iodine, which are more stable, have their absorption peaks at 2900 A for benzene complexes and 3300 A for mesitylene complexes. In general, then, the stronger the complex, the further toward the visible is its complex absorption band. In addition to this absorption band which is characteristic of the complex, we find the absorption spectra of the separate components, although somewhat mddified. For instance, the strong absorption band of iodine at 5200 A is found at 5000 A in benzene -12 com-plex, and at 4900 A in mesitylene - I 2 , the degree of shift of the band being apparently a function of the stability of the complex.^'^ Recently a new absorption band has been found in complexes of benzene with iodine and bromine.30 This band is at 2600 A, and is thought to be a ten-fold intensificat-ion of the 2600 A benzene transition A^g -B 2 u. MOLECULAR ORBITAL THEORT OF MOLECULAR COMPLEXES Having reviewed what is known experimentally about ff electron complexes, we will now examine briefly the current theory concerning these complexes and their spectra. The idea that there is a partial transfer of charge from, say, a benzene molecule B to a complexed bromine mole-cule A, can be expressed mathematically by saying that the wave function for the ground state N of the complex is a l i n -ear combination of a no-bond wave function,f Q(A, B) % f-Q, and a bonding wave function in which there is a covalent bond between the unpaired electrons in A" and B4, ie. f-^ = Y ( A--Bt) w • V u> 7 2 Thus | ^ f a T 0 * Tali w h e r e a »b (D Since f^, Sf Q and are normalized we have a2+ a ab S + b 2 = 1 where S mj<f*Q ^  d r Now Wjg ^  the energy of the complex is given by where Hi,is the exact Hamiltonian operator for the complex. WN may be calculated using second order perturbation theory. From equation 1. we get the secular equation - o now let J f H X ^ M ; J t * H ^ = H „ , j t f t f W, For a sufficiently accurate approximation we may let % = Wo"for a l l D U t the first term, since essentially the 10-difference is small. Now we have WG - WN H o l - S¥Q ,H G - S¥ 0 Wx - Wc or ¥ ^ ¥ 0 -(H o l - S ¥ Q ) 2 (Wj. _ ¥0J ¥ 0, the energy of the no-bond function includes any classical dipole - dipole attractions, as does W^ , which includes as well the attraction energy of the covalent bond. The resonance energy of the ground state is now given by ¥ Q - ¥ N, this being due to interaction between Y0 and Y • wo " WN will be large i f (H o l - S ¥ 0 ) 2 is large, that is, i f y and ¥ overlap strongly, and i f ¥ i - ¥ a is fairly small. If the ground state of a complex is given by equa-tion (1) then there must be an excited state with wave function ^ ' ^  V - b * *d with a a * and b ^ b* Then by the same method as before ¥ E a W-L 4 (H Q L - SWi)2 H i - W0J Thus a transition from ¥ E ^ — ¥ J J should exist, and its transition probability could be calculated. Since a 2 >> b 2 , we see that jr ^  has almost pure no-bond character; while has almost pure ionic character, so that the transition ¥ E < N - ¥ N may be considered an intermolecular charge - transfer process; that i s , light absorption causes an electron to Jump -11-from B to A. This electron transfer band is obviously a property of the complex as a whole, and the absorption spectra of the separate molecules are expected as well, although perhaps slightly modified. It is possible that these intramolecular states may overlap or interfere quantum-mechanically with the charge-transfer excited state, so that i t may be difficult to distinguish the original molecular spectrum. In Fig. 2 may be seen the potential energy curves of the states WN and Wg for the benzene-iodine complex mole-cule as a function of R, the distance between the centres of the two complexing molecules, C Fig. 2. Potential energy curves for the states WM and Wg of the benzene -1^ complex. Ig vert, is the vertical ionization potential of benzene. E vert, is the vertical electron affinity of iodine. This, of course, does not say that complex form-ation does not occur between one component and an excited state of the other, as this very likely happens. The foregoing treatment, due to Mulliken,-^ is the one which has been the most successful in explaining the i n absorption spectra of molecular complexes. -12-PHENOMENA NOT IMMEDIATELY EXPLAINABLE BY CHARGE TRANSFER, There have been phenomena observed, however, which suggest that a charge-transfer transition may be to a non-bonding state of the 1| complex. Kasha has observed a yellow colouration appearing • in solutions of -chloronaphthalene and ethyl iodide which he attributes to the formation of a collisional complex, since the yellow colour disappears in rigid glass. Kasha interprets the yellow colouration as being due to an in-crease in the T <- S absorption in <-chloronaphthalene. This increase in absorption for a forbidden transition is considered to be due to a collisional perturbation which affects the T <- S transition probability.^ Somewhat similar were the conclusions of Reid con-cerning hydrocarbon complexes with s-trinitrobenzene. He found that the complexes, in rigid solution, when irradiat-ed in an absorbing region emit a short-lived radiation in the region of the long-lived hydrocarbon triplet emission. This suggested that the absorption was to a repulsive state of the complex, and that the hydrocarbon is left in its fir s t triplet state,.otheUs-trinltro-benzene in its ground u state. The varying field produced by the decomposition of the complex is assumed to be sufficient to break down the T -> S selection rule, so that the transition will be allowed and therefore short lived.5^,59 13-PHENOMENA EXPLAINED BY COMPLEX FORMATION; The formation of molecular complexes makes possible the explanation of a number of phenomena, among which are energy transfer and fluorescence quenching. Consider f i r s t energy transfer: Foster has found instances of energy transfer between various large organic molecules, for instance between tripaflavine and rhodamine B, in fairly concentrated solutions. He has also developed a theory showing that a long-range dipole-dipole interaction should exist between such molecules. However, working with far more dilute rigid solu-tions of hydrocarbons, Moodie and Reid have shown that long distance transfer of energy is not possible under these con-ditions between randomly distributed molecules of different compounds, but that energy transfer requires complex form-ation. Furthermore, they have found no evidence for com-plex formation between single pairs of molecules, but only between molecules of one hydrocarbon and microcrystals of another. In such complexes, presumably, the microcrystal would act as the "Lewis base", the hydrocarbon molecule as the acceptor. This phenomen explains why i t is soYextremely difficult to obtain some hydrocarbons, such as anthracene and chrysene, free from other hydrocarbon contaminants, such as naphthacene.^ That fluorescence quenching is often due to mole-cular complex formation has been found by a number of invest-igators. 33,34 Bowen and co-workers, for instance, have -14-found that substituted benzenes and toluenes, in quenching anthracene fluorescence in solution, either form true com-plexes with the anthracene or else form loose collisional Q 1 1 T O complexes.7' Complex formation is also indicated as being responsible for quenching of anthracene fluorescence by O 2 and S O 2 in the vapour phase.^ Whether two molecules will form a collisional com-plex or a true complex, the latter being characterized by a negative temperature coefficient, is often determined by steric effects,52,60^ a g w e n a s ^y the tendency of the two components to act as Lewis acids or bases. For instance quenching of rubrene fluorescence by nitrobenzene arises from true complex formation, while quenching by m-dinitro-and s-trinitrobenzene is due to collisional quenching. This seems to be the opposite of what would be expected from previous considerations. However, i f complex form-ation requires that the ring systems be parallel, then i t may be imagined that the nitro groups on the m-dinitro- and s-trinitrobenzenes would interfere with the phenyl groups 13 at right angles to the plane of the rubrene molecule. Also from Bowen comes the information that in crystals of anthracene, fluorescence is quenched by im-purities of acridine or phenazine.^^ This is probably due to the same type of energy transfer as occurs between anthracene and naphthacene impurity,except that here, instead of anthracene excitation energy being emitted by -15 the impurity, as is the case with naphthacene, the energy when transferred to the heterocyclics, is lost by vibration-al deactivation before i t can be emitted. More recently, however, Bowen, among others, '»"'» has supported the idea of quenching being due to long-range energy transfer and long-range /^electron interaction, in certain cases, rather than complex formation.-^ REASON FOR THE PRESENT INVESTIGATION The present investigation of complex formation with heterocyclic compounds was undertaken for several reasons, the most important of which are, f i r s t , heterocyclics are known to form stable complexes with picric acid and trinitro-benzene, which should have interesting spectroscopic propert-ies; and second, heterocyclics might reasonably be expected to form more stable complexes with hydrocarbons than do other hydrocarbons, and might thus be detected in solution with-out microcrystals Of one component. This seems even more possible since Terenin and Karyakin have detected energy transfer between naphthalene and acridine in the vapour phase. 6 5' 6 6' A REVIEW OF THE ELECTRONIC PROBERTIESOOF HETEROCYGLIGS It will be well to review what is known of the energy levels and other electronic properties of hetero-cyclics, as this will show that for heterocyclics, unlike hydrocarbons, even fairly accurate determination of energy levels is/ylt impossible. As was seen in the discussion of the benzene-iodine complex, i t is necessary to be able to determine the energy of a given molecular orbital as well as its symmetry in order to know whether or not i t will enter into stable complex formation. The energy levels of any cata-condensed hydrocarbon may be calculated fairly accurately as is shown by the good agreement obtained between experimental and calculated resonance energies and allowed transition energies. '^', ' ' 'Thus, i f the "acid" molecule may also be treated theoretically-with some accuracy, then i t should be possible to calculate at least relative interaction energies for hydrocarbon complexes, provided simplifying assumptions may be made about the symmetry of the complex, which of course, is not always the case. In contrast with aromatic hydrocarbon^, hetero-cyclics have proven rather difficult to deal with theoretic-ally. The main reason for this is that the diHgonal or coulomb terms in the secular equation (written in matrix form) which correspond to the attraction of a nucleus for its own electron, must vary in heterocyclics; the only way these variations in coulomb integral can be taken into account is by introducing empirical parameters. This method gives quite satisfactory results for electron densities and bond orders, but very poor results for the 40.43 resonance energy.^ ' A more convenient method has been developed of treating an aza-substitution as a perturbation -17' of the parent hydrocarbon system; this method, howe&er, gives 20-23,41 no information concerning energy levels, A semi-empirical method has been devised by Dewar for obtaining the energies of molecular orbitals of pyrrole-type compounds.2£ In pyrrole the G-N bonds have lower bond orders than the C-G bonds, and the nitrogen has more than its share of the six electrons, both facts being due to the higher electronegativity of nitrogen. However, this charge distribution should also exist in the hypothetical ' where the G-X exchange integral is much less than that of C-C, but the coulomb integral for X is equal to that for G. This method gives fairly good values of the resonance energy In pyrrole-type compounds, but, of course, cannot be extended indescriminately to other types of hetero-cyclics. TRANSITIONS FOUND IN HETEROCYCLICS, We will now consider the transitions which are expected and found experimentally in most heterocyclics. ^ It will be convenient to consider the following diagram, Fig.3. G is the ground state, S^  and S 2 the fi r s t and second excited singlet states, and TT_ and T2 are the f i r s t and second triplet states. T^ and T 2 correspond: to the impairing of the excited electron -18-and the one left behind im the now half empty orbital* The solid vertical lines indicate optical transitions, while the dotted ones indicate radiationless transitions. . The trans-ition S + G corresponds to the excitation of an electron from the f i l l e d TT molecular orbital having the highest potential energy, to the lowest antibonding orbital having the correct symmetry to f u l f i l the requirement that the integral for the transition probability, ^s^* where X stands for any distance coordinate, shall remain in-variant under any symmetry operation of the molecule. The optical transition T^ •» G, and the radia-tionless transitions S-^  Ti, etc, are forbidden, since they involve a change in multiplicity. However, these transitions may be made partially allowed by spin-orbit coupling, a phenomenon in which, instead of the ordinary L-S coupling that mainly occurs with small atoms, there is also a small amount of j - j coupling.^»^5 This results in a slight mixing of certain of the singlet levels with the triplet level in question, making transitions between i t and singlet levels partially allowed. Since the amount of spin orbit coupling in a molecule is a function of the square of the charge on the nucleus of the heaviest atom present, heterocyclics should show slightly more allowed triplets than hydrocarbons. Another transition which is expected to some extent in heterocyclics, is the so-called n - *7T transition, in which one of the non-bonding electrons on the nitrogen 19< is excited up to the ff system. This transition is forbidden by symmetry, but has been observed weakly in absorption in pyridine and pyrazine as well as other compounds. 5 7» This transition is recognizable by the fact that i t shows a blue shift in increasingly polar solvents,^-6, and disappears completely upon salt formation, with, say, sulphuric acid. For any transitions which are semi-forbidden to show up in emission, i t is necessary to use low temperat-ures and rigid glass solvents, since i f collisions are possible, deactivation of the metastable state will occur by collision before radiation is possible. -20--;2L-Experlmental Methods and Results. APPARATUS A photograph of the apparatus used in this investir gation will be seen on the proceeding page. The apparatus for studying emission spectra consisted of a General Electric A. H. 6 high pressure, mercury arc with suitable filters as a light source, a Hilger E2 spectograph, and Hilger scanning unit and photographic recorder. Absorption spectra were obtained in some instances using the spectrograph with a Beckman hydrogen lamp, in which case the spectra were photographed on plates. A Beckman spectro-photometer was also employed in obtaining certain of the absorption spectra. Measurements of relative intensit-ies were made using a Hilger non-recording photo-electric microphotometer. A simple phosphoroscope was also used to some ex-tent. This consisted of a metal can with a s l i t in the side which was rotated by means of a small motor. PROCEDURE Solutions of the heterocyclics and their polynitro complexes were made using solvents which will form a rigid glass at -180°C. the temperature of liquid nitrogen. Test tubes of these solutions were frozen down in a transparent quartz or Vycor dewar vessel, and irradiated with 31O0A, 3650 A, or 4353 A light from mercury arc. The emissions -22-were recorded with the scanning unit i n the region from approximately 3500 to 5500 A\. In the region from 5000A up, Eastman Kodak IIF 3 plates were used. The required mercury line was isolated using the following f i l t e r s : Corning f i l t e r 754 with nickel sulphate solution for 3100A, f i l t e r number 737 for 3600A and f i l t e r number 5.74 for 435SA. For absorption spectra below 3 0 0 0 A , the problem of suitable holders for the ri g i d solutions arose, since pyrex 1 does not transmit appreciably below this wave length. A reasonably satisfactory solution to the problem was found in the "popsicle" technique. For this technique a small brass v e s s i l about 10 cm. deep and 1.5 x 1.5 cm. across i s made and f i t t e d with a long handle. The liquid to be frozen i s poured into this vessel, and the whole thing placed in liquid nitrogen. When i t i s hard, a thin rod i s thrust well into the frozen solvent. When the rod i s solidly frozen i n , the whole vessel i s placed i n cold water, and the rod i s used to pull out the block of solvent, the sides of which have now melted s l i g h t l y . This "popsicle" may then be transfer-ed to the quartz dewar, and has the advantage of having f l a t surfaces, a very desirable feature in obtaining absorption spectra. The absorption spectra were recorded on Eastman Kodak IN plates, intensity comparisons being made with pure solvent popsicles at varying exposure times. , The absorption spectra of dry solids were obtained •-23*" with the Beckman spectro-photometer by making a solution of the compound in ether and then evaporating off the ether in a quartz Beckman cell. In the case of pure heterocyclic com-pounds an empty cell was used as a blank, while for the poly-nitro compound complexes, the blank used was a cell with a thin film of the polynitro compound concerned. This technique had certain disadvantages; for instance, in order to obtain accurate readings on the Beckman, the refractive index of the sample and of the blank should be the same; furthermore, the amount of polynitro compound in the film on the blank cell could never be made exactly equal to the amount in the film containing the complex. In spite of these facts, how-ever, the readings obtained were closely reproducible with regard to wave length, and to a lesser extent with regard to relative intensities of the main absorption bands. The glass used for most of this work consisted of twelve parts of ethyl ether, five parts of isopropyl ether, and three parts of n-butyl ether with two parts of isopentane added to prevent the glass from cracking readily. However, the above mixture, although ideal as a solvent for hetero-cyclic compounds, tended to crack upon direct contact with liquid nitrogen. Thus for the "popsicle" technique i t was necessary to use E. P. A., a mixture of five parts of iso-pentane, five parts of ethyl ether and two parts of ethanol. This s olvent was also used to f ind the effect of a more polar solvent on the phenomena observed. - 2 4 * Purification of the chemicals used was an important preliminary to this investigation, although the materials used were a l l reagent grade to begin with. The best way to purify the heterocyclic compounds used was found to be by chromatography through columns of alumina, employing low boiling petroleum ether as solvent. The polynitro compounds were purified by repeated recrystallization, the s-trinitro-benzene and m- and p-dinitrobenzenes from ether, and the dinitrocresols from ethanols. The solvents used had to be both pure and free from water. The isopentane was run through a column containing a layer of silica gel and a layer of alumina, and then distilled. The ethyl ether used was dried with sodium and distilled. The isopropyl and n-butyl ethers were found to be dry, and were just distilled. One hundred percent ethan-ol was distilled. In this way a l l the solvents, when frozen down to -180°C, and irradiated with 3100A showed no emiss-ion, which indicated a complete lack of large unsaturated impurities. RESULTS The polynitrobenzene complexes examined spectroscopy ically during this investigation f a l l into two categories: those containing 2,4-dinitro o^cresol or 2,6-dinitro p-cresol, and those containing one of s-trinitrobenzene, p-dinitro-benzene or m-dinitrobenzene. Complexes of the f i r s t type have some salt-like character, being reasonably soluble in water, a fact which* -25 indicates that they are, to some extent at least, of the form RNH* OC^fCI^) (N02)2~> especially with fairly strong bases such as the quinoline compounds. These complexes dissolved in ether show a large increase of colour intensity upon cooling to liquid nitrogen temperature. On the other hand, complexes with the di- and t r i -nitrobenzenes behave as non-ionic dispersion force complexes as would be expected; they are very insoluble in water. Their absorption intensities indicate a large negative temp-erature coefficient. A complex formation between a heterocyclic compound and a nitro compound could usually be recognized at once by the appearance of colour, generally yellow, although in very dilute solutions this colour would not appear until the complex is cooled down considerably. Of the heterocyclic compounds studied, only acridine and the dibenzacridines absorb as high as 3600 A. The other heterocyclics a l l have their longest wave length absorption band at 3100 A or a bit lower. The nitro complexes, on the other hand, absorb up into the blue. The emission spectra of the complexes, other than the acridine complexes, were examined with irradiation of 3100 A. and 3650 A. In investigating the complex emissions using 3100A excitation, varying amounts of the complexing nitro compound were used with constant amounts of heterocyclic V 26-compound. In many cases in which complex formation occurred, complete quenching of the heterocyclic fluorescence took place upon addition of the non fluorescing nitro compound* This was especially the case with the carbazolfe-s-trinitrobenzene com-plex. However, in other instances, quenching took place more slowly as the amount of nitro compound in solution was increased; furthermore, interesting changes in the emitted radiation could be observed. Consider, for instance, the benzo -(f)- quinoline complex formed with 2,6-dinitro -p^cresol solutions to which have been added equal molar quantities of both components (approximately 1 x 10"^ mols/l.) s t i l l show a long-lived emission in the region of the triplet emission of the benzo (f) quinoline at 4400 to 5300 A. r, but there is l i t t l e band structure remaining. The emission in the region of the 4100 A. '. triplet band is lost by reabsorption; also the singlet emission of the heterocyclic completely dis-appears when there is any dinitro-cresol complex present. A phosphoroscope determination of the emission of this solution was made, so that the long-lived component of the radiation 4 could be isolated, and i t was found that the structurless emission in between the two long-lived, phosphorescence bands is short-lived. (Fig.4. (a) - (e).) However, i t was found possible to add enough dinit©-cresol to the benzo (f) quinoline solution to entirely quench the fluorescence of the heterocyclic compound, and in this case there is no sign of any short-lived emission ether. -27= The addition of s-trinitrobenzene or m- or p-dinitro-benzene often resulted in partial or complete quenching of a heterocyclic emission, although a loss of structure of the triplet emission was not apparent. The triplet band nearest the blue usually disappears when there is nitro complex present, as does the singlet emission. Indole, however, was an exception. Its faint short-lived triplet bands in the yellow green were enhanced slight-ly and made short-lived by complex formation with p-dinitro-benzene. A summary of the results obtained using 3100 A ex-citation will be found in Table 1. Now emission spectra obtained with 3650 A as the ex-citing light will be considered. Acridine absorbs at this wave length, but not much at 3100 A. The acridine-s-trinitro-benzene complex irradiated with 3600 A exhibits short-lived emission in the-acridine triplet region, as has been reported elsewhere.^ Now of the complexes of the other heterocyclics, many emit a fairly intense, short-lived radiation in the region of the triplet bands of the heterocyclic, when irradi-ated with 3600 A light. (Figs. 5, 6, 7). The results obtain-ed for the various complexes are tabulated in Table 31. It is most important to note that a solution of a heterocyclic compound, such as benzo(fJquinoline with enough nitro compound added to completely quench a l l emission with 3100 A irradiation, will emit strongly in the heterocyclic -28~ triplet region when irradiated with 36OO A light. The above results may be obtained both in ether glass and in the more polar E.P.A. The emission spectra of the dry solids of most of the foregoing complexes were investigated. In a l l cases in which a complex in solution showed an enhanced triplet, the dry solid did also, when irradiated with light of the same wave lengtho This emission of the dry solid was usually shifted a bit to the red of the triplet emission of the hetero-cyclic. However, there were two cases in which irradiation of the dry solid complex with 3100 A produced the same short-lived emission a<s did 36OO A light. This occurred "in spite of the fact that the complex in solution did not show a short-lived triplet with 3100 A irradiation. For example, this occurred with benzo(f)quinoline, -2,4-dinitr© o-cresol com-plex, and xanthene-trinitrobenzene complex. It should be noted that other complexes in the dry solid state, such as the dibenzo-P,'uran-s-trinitrobenzene complex, do not emit appreciably with 3100 A irradiation, but do show an enhanced triplet with 3600 A irradiation. The s-trinitrobenzene complexes of 1, 2, 5, 6- and 1, 2, 7, 8-dibenzacridine were examined. Using 4358 and 36OO A exciting light, i t was possible to see visually that the reddish yellow triplets were enhanced and short-lived in 0 the complex, although the free heterocyclic triplet lifetime, when irradiated with 36OO A is quite long-lived. (Fig. 8). It is interesting to note that the T-N-B complex of the -29-former compound is pale yellow, while that of the latter is purple. The dry solids of both complexes show the same red-dish yellow emissions as the solutions, although not so bright. These complexes were rather difficult to study in solution, because they are so very insoluble in ether. The absorption spectra of the complexes trinitro-benzene-xanthene and 2,4-dinitr© o-cresol-benzene(f)quinoline were studied, both in solution and as dry solids, and compar-ed with the absorption of the pure heterocyclic compounds* In both cases i t was found that the JJabsorption of the heterocyclic was retained at the same wavelength in the complex, a continuous absorption region extending from the end of the long wave limit of heterocyclic absorption up to 4500-5000 A. (Figs. 9, 10). It was assumed that the nitro compound absorption remained pretty well unchanged as a decreasing absorption from about 2600 A to 3000 A, so that the blank of nitro compound could be used. The absorption spectrum of benzo(f)quinolinium hydrochloride was also obtain-ed, (Fig.10), and i t will be seen that there is an extra absorption peak at 3650 A which appears neither in the free base, nor in the complex.^2 In an effort to determine whether any of the long-lived emission in benzo(f)quinoline is w-^JJ. the emission was observed using the increasingly polar series of solvents, a. Four parts methyl cyclohexane plus one part isopentanej b. Ether mixture (see page 23)\ c. E. P. A.j d. eight parts ethanol, one part methanol, one part isopropanol. How--30-ever, no detectable shift was observed. It was thought that AlCT^, being a good Lewis acid, might form molecular complexes with heterocyclics. However, its effect was observed with quinoline and carbazole, and there was no sign of ^ complex formation. This is not sur-prising since A I C I 3 does not form JJ-complexes with hydro-carbons.^ Since-trimesic acid has an electron deficient ring, i t was thought that i t might form complexes with heterocyclics, and indeed, i t does form a yellow complex with acridine, although not visibly with any of the other heterocyclics test-ed. However, neither acridine-trimesic acid solutions nor those with any other heterocyclic compounds showed any emissions that were not identical with the heterocyclic emissions. The heterocyclic compounds were a l l examined for complex formation with naphthacene solution and suspension. But there was no sign of complexes with naphthacene mole-cules, although there were three instances of complexes between heterocyclic and naphthacene microcrystal. These include dibenzofuran and benzofuran, both of which show a new band at 5250, indicating energy transfer to naphthacene suspension,^® and dibenzothidptoene which shows quenching of emission in contact with naphthacene suspension. Table 1. Effect on heterocyclic triplet emission of complex form-ation, in rigid solution, irradiating with 3100 A. '. 2,6-dinitro-p-cresol 2,4-dinitro-o-cresol s-trinitro-benzene p-dinitro-benzene Indole Carbazole Benzofuran Dibenzofuran Dibenzothiophene Xanthene Quinoline (also quinaldene) Ben^o(f)quinoline (also .^methyl benzo(f)quinoline quenching no apparent complex. no apparent complex. some quenching. some quenching* quenching quenching (also dry solids^ (also dry solids) quenching quenching no apparent complex. no apparent complex. loss of structure and quench-ing. loss of structure and quench-ing. quenching quenching(also for dry solids). seme quenching. quenching falsO dry solids) some quenching, some loss of band structure. enhanced trip-let , 4600-6000A1-quenching felso dry solids) quenching not completely complexed, some quenching, some enhanced triplet (more ehhanced triplet with dry solid). quenching. May not complex. Table 11. Emission with exciting light of 36OO A (in rigid solution) 2^dInitror p-ereaol 2,4-dinitro-o-cresol s-trinitro-benzene p-dinitro-benzene Indole Carbazole Benzofuran Diobenzofuran Dibenzothiophene Xanthene Quinoline (also quinaldene) Benzo(f)quinoline (also 3-mefahyl benzo(f)quinoline) faint short-lived, yellow. faint short-lived, 4600 - 6000A no emission. no emission. no emission, short-lived emission in triplet region - - ). strong short lived, 4600 - 6000 A. no emission. short-lived emission in triplet region. ehhanced short-lived emission in triplet region. short-lived emission in triplet region - - - - - - - - - - -quenching. - - ) . strong short-lived emission in triplet region. strong short-lived emission in triplet region. short-lived emission in triplet region. faint emission Faint emission in in triplet region, triplet region. quenching• - 3 3 -) \ a* b. ^^^^^^^^ c. ' d. / 1 \ S — 1 1 ' - L _ l 4500 5000 5500 4500 5000 5500 4" 500 5600 5^00 Wave length in Angstrom units. F i g . 4 ; Emission spectrum of benzo(f)quinoline, and of the benzo(f)quinoline - 2 , 6-dinitro p-cresol complex,-i n solution i n ethers, i r r a d i a t e d with 3100 A; A l l the emission spectra shown here ere long-lived, a. be"nzo( f) quinoline; b.. benzo(f)quinoline, 2 , 4 -d i n i t r o p-cresol, r e l a t i v e molar quantities 2 :1 ; c. same as b; proportion 1:1; d. as b, except proportion 1:2; e. Same as f o r c, using phosphor-scope. -34-Wave length in Angstrom units. Fig, 5. Emission spectrum of dibenzofuran, and of i t s complex with s-trinitrobenzene. a. dibenxofuran irradiated with 3100 A, long-lived. b. dibenzofuran-s-trinitrobenzene complex irradiated with 3609 A, short-lived. Wave length in Angstrom units. F i g . 6 . Emission spectrum of xanthene, and of i t s complex with s-trinitrobenzene. a. xanthene ir r a d i a t e d with 3100 A, long-lived component. b. Solid curve, xanthene s-ftinitrobenzene complex irradiated with 36OO A, sh o r t - l i v e d . Dotted;-curve - complex dry s o l i d with 36OO A l i g h t . - 3 6 -4500 5000 5500 Wave length in Angstrom units F i g . 1. Emission spectrum of benzo(f)quinoline, and of i t s complex with 2,4-dinitro o-cresol. benzo(f)quinoline i r r a d i a t e d by 3100 A, long-lived portion. benzo(f)quinoline - 2,4-dinitro o-cresol complex irradiated with 3600 A, shhrt-lived. c. Dry s o l i d complex i r r a d i a t e d with 3600 A. a. b. - 3 7 -a. Wave length i n Angstrom units. 4 F i g . 3. Emission spectrum of 1, 2, 7, 8- dibenz-acridine and of i t s s-trinitrobenzene complex. a. 1, 2, 7, 3-dibenzacridine alone, long-lived emission. ^ N b. s-trinitrobenzene complex in r i g i d ether solution, short-lived. -38-o 2500 2?50" 3000 3?50 JW Wave length in Angstrom units. Fig. 9 . Absorption spectrum of xanthene and of i t s complex with s-trinitrobenzene, both as dry solids. a. xanthene; b. xanthene-s-trinitrobenzene complex -39-CD O o 3000 3200 3400 3600 3800 Wave length in Angstrom units. Fig.10. Absorption spectrum of benzo(f)quinoline and of i t s complex with 2,4-dinitro-o-cresol. a . benzo(f) quinoline alone in rigid E.P.A.- b. 2,4-dinitre-o-cresol complex in E.P.A.; c. benzo(fjquinoliniura hydrochloride in ethanol at room temperature. -39a-DISCUSSION First we will consider the phenomena which take place with benzo(f)quinoline-dinitro-cresol complexes. As was mentioned earlier, a solution may be made so concentrated in nitro compound that i t is completely quenched with 3100 A radiation, and yet show strong fluorescence in the triplet region of the benzo(f)quinoline when irradiated with 3600 A light. We have seen from the absorption spectra diagram (Fig. 10) that the complex absorbs at 3100 A even more strongly than i t does at 36OO A. The occurrence of strong emission upon irradiation with 3600 A and not 3100 A must mean that in the two cases absorption is to different excited states. Now i f the absorption region which lies at longer wave lengths than the original heterocyclic // absorption is a charge-transfer transition directly to a non-bonding state of the complex, such that the complex dissociates leaving the heterocyclic compound in its triplet state, then one would expect that irradiating the complex with a slightly higher energy radi-ation would produce the same results. Thus i t seems likely that the transition excited with 31Q0 A. is to a bonding state of the complex. That i s , the state c^c-nip" s A^ 4 Y (B+- A") can be a bonding state of the complex as long as the symmetries of the orbitals are correct, where ^ represents the fi r s t excited singlet of the heterocyclic compound belong-ing, to some extent, to the whole complex. From here the ex-citation energy is lost by vibrational deactivation and radia--40-ationless crossings to the ground state. One would not ex-pect the complex as a whole to emit this energy in an optical transition in view of the presence of vibrating nitro groups and other factors. It seems likely that the charge-transfer transition excited by 36OO A light is to a bonding state of the complex also, namely to the state E described on page 1 1 . However, this state E is at very closely the same energy as the hetero-cyclic triplet level, and the possibility of a radiationless crossing to the heterocyclic triplet level is suggested. If such a transition state should occur, then the complex would be in a repulsive state and would dissociate. The changing field produced by dissociation could be assumed to make the transition to the ground state allowed, as was mentioned on page 1 2 . There are two apparently serious objections to the idea of such a radiationless transition: these being that the transition would be both multiplicity and Frank-Condon forbid-den. Dealing with the multiplicity selection rule f i r s t , i t seems that the objection is not too serious since spin-orbit coupling will occur, to produce mixing into the triplet state of some singlet character and vice versa. This effect will be quite small since the N and 0 atoms present have quite small Z values. However, spin-orbit coupling is sufficient in other compounds such as hydrocarbons to produce a radia-tionless transition from excited singlet to triplet so that this selection rule will nefctainly not be unsuperable here. Turning now to the objection thati:the inter-combin-ation is Frank-Gondon forbidden,it will be convenient to con-sider the complex as two potential wells separated by a fairly high barrier. This can be looked upon as an extension of the free electron method, ^ »^» 61» similar to the extension made by 62 Shuler for the case of a two dimensional TT electron complex. Wave functions will look more complicated for actual IT complex bonding orbitals, since their electr-onic quantum numbers will be high; however, the tunnel effect will occur even more for higher orbitals. Fig.11. Two different potential wells separated by fairly high barrier. The barrier height is proportional to - Ig g A being the electron affinity of the Lewis acid, and Ig the ionization potential of the Lewis base. Now i f we take 1 as the perturbed ground state level of the heterocyclic, then 1 to 2 may be con-sidered the charge-transfer transition; while 1 to 3 will be the slightly modified transition to the fir s t excited singlet of the heterocyclic compound. Now suppose that the ^biplet state corresponding to 3 is below 2 in energy. The statement that the transition 2 to 3 (triplet) i s Frank-Gondon forbidden means that an electron cannot make the transition without a -42-change in its position. At first one would tend to think of the transition from a charge-transfer state to the heterocyclic triplet as a movement &f the electron across the complex to some extent. However, observing Fig. 11, one can see that the electron, even in its excited state 2, has a considerable possibility of being on the heterocyclic molecule. Thus i t should be possible for a 2 to 3(triplet) transition to take place after a l l , on these considerations at least. If spin-orbit coupling in the complex is responsible for the multiplicity selection rule being broken down somewhat, then there arises the question of why the complex,when in the excited but modified heterocyclic singlet state, does not show some radiationless transition to the triplet for the same reason of enhanced spin-orbit coupling. However, i f the effect of triplet mixing into singlet states is to be con-sidered as well as the reverse, then i t must be the case that the level 2 will be more affected by spin-orbit coupling than level 3 (singlet). Once an orbital has become perturbed by spin-orbit coupling then its effect would have to be much the same over the whole complex for that orbital, since an elect-ron cannot be expected to change its degree of singlet or triplet character according to its position in an orbital. However, the nitro groups in the "acid" part of the complex are responsible for the strongest spin-orbit coupling, and surely the orbital with the highest probability density on the nitro compound will be the most affected by spin-orbit coupling. - 4 3 -In addition the charge-transfer excited electron would not have the same probability of deactivation to ground as would the fi r s t excited singlet of the heterocyclic, as i t involves very different transitions back to ground, so that i t is not essential that radiationless transitions occur to the same degree for both levels, even i f the spin-orbit coupling argument did not exist. The energy of the charge-transfer transition can be expected to increase with decreasing strength of complex formation. This will provide an explanation of why the dry solids of certain complexes, eg., dinitrocresol and benzo(f) quinoline, emitted a short-lived triplet radiation with 3100 A light when the solution did not. The forces between mole-cules of a solid are not enough to change the wave length of absorption of an ordinary molecule, although they are respons-ible often for small maxima beyond the long-wave length end of the absorption of the molecule. However, in the case of weak complexes, the forces between molecules of complex will be of almost the same order as the forces between the components of the complex, and a shift of the charge-transfer spectrum to higher energies is expected. Another way of looking at this is that the attractive forces between complex molecules in a crystal provide-, a lowering of potential energy for the ground state. This will result in a blue shift of any transition to a • state differing a great deal from the ground state. It seems likely that the benzo(f)quinoline forms a slightly fluorescing complex with trinitrobenzene because of -44-the increased strength of the complex, so that the excited charge-transfer state will not be much above the triplet state of the heterocyclic, and the cross-over will occur only to a small extent. The tendency of the dinitrocresols to form true salts in polar solvents will not increase the tendency for /Jcomplex formation, rather the added hydroxy group will activate the ring and decrease the tendency for the ring system to act as a Lewis acid. The fact that the complexes of benzo(f)quinoline-dinitrocresol showed short-lived emission between the triplet bands of the heterocyclic with 3100 A irradiation must be due to the reabsorption by the complex of strong heterocyclic singlet emission. The fact that indole shows an enhanced triplet upon irradiating with 3100 A light is probably due to the fact that i t forms a weaker complex than many of the other hetero-cyclics. Although the carbon atoms on theipyrrolejring have a higher charge density than they would in, say, benzene, yet the [J orbitals are not as free to yield electron density in further bond formation because of the competition for electrons which comes from the presence of electronegative nitrogen. These considerations will result in a slight less-ening of aromatic character for the indole ring system, and a lowering of polarizability. This lessening of aromaticity resulting from not having completely delocalized /^"orbitals will result in a lowering of the energies of the top bonding orbitals, so that a charge-transfer transition will be of =45-' higher energy than for a more fully aromatic ring system. The same consideration will be true, though to a lesser extent, for carbazole. However, the reason carbazole complexes do not fluoresce with irradiation of either 3100 A or 3600 A is probably that the carbazole triplet levels are eg at too high energies. Xanthene, too, should form weaker complexes and its complexes have a higher energy charge-transfer transition because of the lessening of aromaticity resulting because the conjugation of the ring system is broken to some extent. The quinoline and acridine complexes, on the other hand, should be stronger, since there is no question of less-ening of aromatic character and resulting loss of polariz-ability. The fact that trimesic acid did not result in complex formation other than with acridine is not surprising, since i t will be a much weaker Lewis acid than is a polynitrobenzene compound. Even i f i t did form complexes, the resulting absorp-tion region would not be expected to be in the same region as for the nitro complexes, so that cross-over to the hetero-cyclic triplet would not necessarily be expected. It is also not too surprising that complex formation was not observed between heterocyclics and hydrocarbon mole-cules, since there is not enough difference between the elect-ron densities on the two rings. In the case of hydrocarbon crystals, however, the matter is quite different, since elect-rons are known to move with a certain degree of freedom throughout the whole crystal, and the electron density at the - 4 6 -surface of the crystal could be expected tobe high, so that the microcrystal could act readily as a Lewis base, in the presence of heterocyclic molecules, which will be capable of accepting some of this aumplus electron density, especially since in benzo and dibenzofuran^the compounds for which trans-fer to hydrocarbon crystals occurred, the furan ring system has not got the complete six electrons i t needs for aromatic character in a state delocalized from the oxygen atom* Further work on the problem of heterocyclic com-plex formation with nitro compounds should include the follow-ing investigations: 1) . More detailed examination of the absorption spectra of the complexes at low temperatures with equipment giving higher resolution that i t was possible to obtain here. 2) . Investigation of exact quantum yields of these "complexes, with a view to finding how much energy is lost by vibrational deactivation and other means. 3) . Polarization studies coupled with theoretical investigations of the expected polarizations of the various theoretically expected trans-itions o =47-' - BIBLIOGRAPHY -1. S.L.Altman, Proc. Phys. Soc. (London) A63_, 1234(1952). 2. L.J.Andrews and R.M.Keefer, J.Am. Chem.Soc. 22., 462(1951). 3. L.J.Andrews and R.M.Keefer, J.Am.Chem.Soc. 2i, 4169(1951). 4 . N.S.Bayliss, Quart. Rev. 6 , No.4, 319(1952). 5. H.A.Benesi and J.H.Hildebrand, J.Am.Chem.Soc.20, 2832(1948). 6 . H.A.Benesi and J.H.Hildebrand, J.Am.Chem.Soc.21,2703(1949). 7. G.Berthier and B.Pullman, Compt.Rendu. 23.1, 774(1950). 8. N.W.Blake, H.Winston and J.A.Patterson, J.Am.Ghem.Soc. 22, 4437(1951). 9. E.J.Bowen, Quart.Rev.Ghem.Soc. (London) 1,, 1(1947). 10. E.J.Bowen, Symposia of the Society for Exp'tal. Biology, ix (1951). 11. E.J.Bowen, A.W.Barns and P.Holliday, Trans. Faraday Soc. Ltl. .27(1947). 12. E.J.Bowen and E.Coates, J.Chem. Soc. 105(1947). 13. E.J.Bowen and E.Coates, J.Chem.Soc. 131(1947). 14. E.J.Bowen and P.D.Lawley, Nature, 16/j., 572(1949). 15. E.J.Bowen and W.S.Metcalf, Proc.Roy.Soc.(London) A206. 437(19510. . 16. E.J.Bowen, E.Mikiewicz and F.W.Smith, Proc.Roy.Soc. (London) A62, 26(1949). 17. E.Boyland, Cancer Research, 12,' 77(1952). 18. L.G.S.Brooker and W.T.Simpson, Ann.Rev.Phys.Chem.2^121(1951) 19. H.C.Brown,H.Pearsail and L.P.Eddy,.J.Am.Ghem.Soc.22, 5347(1950). 20. C.A.Coulson and H.C.Longuet-Higgins, Proc.Roy.Soc.(London/ A191. 39(1947). -48-21. C.A.Coulson and H.C.Longuet-Higgins, Proc.Roy.Soc.(London) A192. (1947). 16. 22. C.A.Coulson and H.C.Longuet-Higgins, Proc.Roy.Soc.(London) AlSa,447(1948). 23. C.A.Coulson and H.C.Longuet-Higgins, Proc.Roy.Soc.(London) A1932 456(1948). 24. M.J.S.Dewar, The Electronic Theory of Organic Chemistry (Oxford,at the Clarendon Press,England,1949). 25. M.J.S.Dewar, Trans Faraday Soc. 1+2, 764(1946). 26. K.Fukui, T.Yonezawa and H.Shingu, J.Chem.Phys.20, 722(1952). 27. M.D.Galanin, Izvest. Akad.Nauk., S.S.S.R., Ser. Viz. 1£, 5-43-9-(1951).C.A. 1+6, 8971b(1952). 28.. E.H.Gilmore, G.E.Gibson and D.S.McLure, J.Chem.Phys. 20. 829(1952). 29. F.Halverson and R.C.Hirt, J.Chem.Phys. 12, 711(1952). 30. J.S.Ham, ff.R.Piatt and H.McConnell, J.Chem.Phys. 19, 1301 (1951). 31. D.L.Hammick and R.B.M.Yule, J.Chem.Soc. 1539(1940). *32. J.M.Hearn, R.A.Morton and J.C.E.Simpson, J.Chem.Soc. 3318(1951). 33. A.V.Karyakin, Izvest.Adad.Nauk. S.S.S.R., Ser. Fiz. 15. 556(1951). C.A.-4J>, S971h(1952). 34. A.V.Karyakin, Zhxxri Fiz. Khim. 26, 96U952). CA. 46, 9983^(1952). "~ . Soc. 35. M.Kasha, Disc. Faraday/on Spectroscopy and Molecular Structure, 14(1950). 36. M.Kasha, J. Ghem. Phys. 20, 71(1952). 37. H.B.Klevens and J.R.Piatt, J.Chem.Phys. 12, 470(1949). 38. J.Landauer and H.McConnell, J.Am.Chem.Soc. Zkt 1221(1952). 39. G.H.Locket, J. Chem. Soc. 1501(1932). 40. H.C.Longuet-Higgins and C.A.Coulson, Trans. Faraday Soc. Al, 80(1947). 0 -49-41. H.C.Longuet-Higgins and C.A.Coulson, J.Chem.Soc.971(1949). 42. H.C.Longuet-Higgins and G.W.Wheland, Ann.Rev.Phys.Chem. Vol. 1, (1950)133. 43. P.Lowdin, J. Chem. Phys. 12, 1323(1951). 4 4 . D.S.McClure, J. Chem. Phys. 12, 905(1949). 45. D.S.McClure, J. Chem. Phys. 20, 682(1952). 46. H.McConnell, J.Chem. Phys. 20, 701(1952). 47. H.McConnell, J.S.Ham and J.R.Piatt, J.Chem.Phys.21,66(1953X 48. M.M.Moodie and C.Reid, J.Chem. Phys. 20, 1510(1952). 4 9 . R.S.Mullikin, J. Am. Chem. Soc. 21* 6 0 0 U 9 5 0 ) . 50. R.S.Mullikin, J. Am. Chem. Soc. 2ik> 811(1952). 51. B.S.Neporent, Izvest. Akad. Nauk. S.S.S.R., Ser. Fiz. CA. Uh, 8970N(1952). 15., 533(1951). 52. M.Orchin, J. Org. Chem. 16, 1165(1951). 5 3 . P.Pfeiffer/if Ann. ^^ 12, 291(1916). 54. J.R.Piatt, J. Chem. Phys. 12, 484(1949). 55. J.R.Piatt, J. Chem. Phys. 18, 1168(1950). 56. J.R.Piatt, J. Chem. Phys. 1£, 101(1951). 57. CC.Reid, J. Chem.Phys. 18, 1673(1950). 58. C Reid, J. Chem. Phys. 20, 1212(1952). 59. C Reid; J. Chem. Phys. 20, 1214(1952). 60. J.C.Rowell and V.K.LaMer, 3. Am.Chem.Soc. 22»l630(1951)• 61. K.Rudenberg and R.G.Parr, J.Chem.Phys. 1£, 1268(1951). 6.2. K.E.Shuler, J. Chem. Phys. 20, 1865(1952). 6 3 . H.Shull and F.O.Ellison, J.Chem.Phys. lg, 1610(1951). 64. H.Sponer and E.Teller, Rev. Mod. Phys. 13_, 99(1941). -50-6 5 . A.N.Terenin and V.L.Ermolaev, Doklady Akad. Nauk. S.S.S.R., |£, 547(1952). CA. A6, 9994b(1952). 6 6 . A.N.Terenin and A.V.Karyakin, Doklady Akad. Nauk. S.S.S.R., jtf, 425(1947). CA.A6, 4373a(1952). 67. A.N.Ternin and A.V.Karyakin, Izvest. Adad. Nauk. S.S.S.R., Ser. Fiz. lj>, 550(1951). CA.A6, 8981e(1951). 68. J.Weiss, J.Chem. Soc. 245(1942). 69. R.B.Woodward, J.Am. Chem. Soc. 64_, 3058(1942). 51-Appendix. Syntheses of heterocyclics undertaken. BENZOFURAN1 (a) Preparation of o-formyl phenoxy acetic acid, The interaction between o-salicylaldehyde (5 g.) and ethyl bromo acetate (6 g.) took place in boiling acetone (60 cc) containing potassium carbonate (10 g.) over a period of three hours, o-formyl "phenoxyaeetic ester was thus obtained as colourless prisms melting at 47-48°C. Hydrolysis of the ester with 4$ aqueous sodium,hydr-oxide for ten minutes gave the desired acid, which was obtain-ed as colourless needles, melting at 132°C, after recrystall-iaation from benzene. (b) Preparation of benzofuran, A mixture of the above acid (ten parts by weight) and sodium acetate (28 parts) with acetic anhydride (100 parts) were boiled for one half hour. The mixture was dilut-ed with water and neutralized the next day with sodium bi-carbonate. The benzofuran was isolated with ether, and was an almost colourless mobile o i l boiling at 132°C. 1, 2, 7, 8 DIBENZACRIDINE3 /3 -napthol (5 g.) and #-naphthylamine (5 g.) were melted together at 160°C. Paraformaldehyde (1 .5 g.) was gradually added. The latter quickly dissolved with the evolution of water vapour, and the melt became yellow. The mixture was heated another hour at 200-250°G. =52 then i t was poured into warm, dilute sodium hydroxide solu-tion. The mixture in the sodium hydroxide solution was boil-ed up. The dibenzaaiidine then settled as a yellow granular; mass, which w%s filtered, washed and dried. By recrystall-izing this yellow substance several times from toluene, bright yellow needle-like crystals of the dibenzacridine (melting at 2l6°C) were obtained in good yield. 1, 2, 5, 6, DIBENZACRIDINE^-The same method was used as for 1, 2, 7, 8 dibenz-acridine exeept that «C-naphthylamine was condensed with p naphthol. In this case the yield was rather low, as the therm-al condensation resulted in a great deal of tarry material being formed. The product obtained was recrystallized from tolu-ene, then several times from acetone, giving bright yellow crystals(melting at 228°G). DIBENZOTHIOPHENE2-Biphenyl (50 g.) and sulphur (21 g.) were melted to-gether in a one l i t r e round bottomed flask, and the temp-erature raised to 115°C at the end of the eighth hour. When cooled, the mass in the reaction flask was extracted three times by boiling with 100 cc's of water, then i t was cooled and the water decanted each time. Then eight alcohol ex-tractions were made, boiling each time with a l i t r e of alcohol and decanting while hot. The combined extracts were digested with darcojsarbon and filtered immediately through a hot Biichner funnel. Upon cooling, almost colourless needles separated from the f i l -trate. These crystals were recrystallized from ethanol and ehromat©graphed(melting at 9 8 ° ) . The compound forms a picrate which melts at 125° C. BIBLIOGRAPHY FOR SYNTHESES 1*. Foster and Robertson, J. Chem.Soc, 921(1939). i* Gilman and Jacoby, J. Org. Chem., 108(1938). j. Ullmann and Fetvadgian, Ber. _}6, 1027(1903). 


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