POLYMETHYLENE PYRIDINES

, A thesis submitted by

David Robert Brittain

in partial fulfilment of the

requirements for the degree of

DOCTOR OP PHILOSOPHY

in the University of London

Organic Chemistry Department, dune, 1965. Imperial College, LONDON, S.W.7. ABSTRACT

This thesis describes a series of attempts to syn- thesise 2,5- and 1,4-polymethylene bridged pyridines. Nuclear magnetic resonance theory predicts that protons, which are held directly over an aromatic ring, will be abnormally shielded compared with protons in aliphatic straight-chain hydrocarbons. This prediction has been verified for the central methylene protons of paracyclo- phanes. The degree of shielding, expressed in terms of the distance from the aromatic ring, is a measure of the induced ring current and hence the aromaticity of the benzene ring. Similar measurements upon 2,5- or 1,4— polymethylene bridged pyridines would make it possible to determine the degree of aromaticity of the pyridine ring relative to benzene. A review of the subject of aromaticity is presented in which special reference has been made to its inter- pretation by nuclear magnetic resonance. The synthetic work has not been brougL.t to a truly satisfactory conclusion. However, the synthetic routes to 2,5-dialkylpyridines have been thoroughly investigated and a wide variety of such compounds prepared. The functional groups at the ends of the alkyl chains have been varied in an effort to produce a derivative which would cyclise to give a 2,5-bridged pyridine. The attempted intramolecular oxidative coupling of 2,5-dihex- 51 -ynylpyridine received much attention. In the attempts to obtain a 1,4-bridged pyridine, two tricyclic compounds, each containing two quaternised pyridine rings linked by polymethylene chains, were obtained. As a result of these investigations, several alter- native routes to bridged pyridines are suggested. ACKNOWLEDGEMENTS

This work was carried out under the supervision of Dr. J. A. Elvidge, D.Sc., to whom I am most grateful for his constant and generous help. I am also indebted to Miss J. Cuckney and her staff for microanalyses and to Mrs. I. Boston for nuclear magnetic resonance measurements. Thanks are also due to Mx. Watson for constructing a microaddition apparatus and for his continual help in the laboratory. I wish to thank the Scientific Research Council for the award of a maintenance allowance. INDEX Page CHAPTER 1: Aromaticity 8 CHAPTER 2: Nuclear Magnetic Resonance and Aromaticity 19 CHAPTER 3: The Synthesis. Introduction 32 Methods for the Synthesis of Macrocyclic Compounds 34 2252111gLaLIZE1111ta: Scope of Work 46 Attempted Synthesis of 5-101-formyl- decy1-2-methylpyridine 47 Ziegler's Method: Pyridine derivatives 50 Cyclopentadecanone 62 Diacetylene Method 64 Preparation of Compounds for Synthetic. Roues a)Pyridine Derivatives 70 b)Aliphatic halides 73 LA2ziclEtdDGIcilaua1.241 74 Addition of Hydrogen bromide to Olefins 80 CHAPTER 4: Speculative Methods for the Synthesis of Bridged Pyridines 83 6 22ga CHAPTER 5: Experimental 89 2:5 Bridged Pyridine 91 1:4 Bridged Pyridine 13V References 148 7

TO MIRIAM - 8

CHAPTER 1

AROMATICITY

Introduction. Benzene, the original aromatic compound, was first isolated by Faraday in 1825. Benzene combines a high degree of apparent unsaturation (from its molecular formula, 06H6) with a high thermodynamic stability and with a reluctance to undergo addition reactions. The conditions required to persuade benzene to undergo addition reactions, are. uch more vigorous than those required for similar reactions of olefinic compounds. Catalytic hydrogenation of benzene requires a combination of high pressure and an elevated temperature. Addition of halogens to benzene requires photochemical activation of the benzene molecule. The characteristic reaction of benzene is one of substitution and even those substitution reactions are not easily accomplished. The task of correlating these chemical and thermo- dynamic properties of benzene with the structure of the benzene molecule, begun by Kekule (1) in 1865, did not gain momentum until the nineteen-twenties. Long before this, a concept of an Ihromaticc, compound had emerged in terms of chemical reactivity, and, embraced by this concept, were the aromatic heterocycles and polynuclear benzenoid hydrocarbons, as well as benzene itself. As long ago as 1880, Thomsen (2) had observed that the heat of combustion of benzene was incompatible with the existence of three discrete double bonds in the molecule. In the attempts to explain aromatic character, the change of emphasis from the study of chemical reactivity to concern with physical properties brought considerable success. There emerged from the work of HIckel, Pauling and their co-workers (3,4) two approximately quantitative descriptions of aromatic molecules. These two descriptions, which are based on the Valence Bond theory and the Mole- cular Orbital theory respectively, are expounded in standard works (5,6). It is upon the success of these descriptions that the present understanding of aromaticity is based.

Resonance Energy. a) Theoretical Calculation. Experimental heat of combustion data show that benzene is about 36 kcal./mola more stable, thermodynamically, than the hypothetical cyclohexatriene. This additional stability is known as the resonance energy of benzene. - 10 -

Broadly speaking, compounds with ,a similar additional stability fall within the older concept of aromaticity, which is expressed in terms of chemical reactivity. Both theoretical descriptions of aromatic molecules allow the calculation of resonance energies. The resonance energies are expressed in terms of an energy parameter which is known as the resonance integral (0) in the M.O. description and the exchange integral (J) in the V.B. description. The total %-electron energy of an aromatic molecule requires a second energy parameter to describe it fully. This is known as the coulomb integral and is denoted by a and Q in the M.O. and the V.B. descriptions respecttvely. For benzene, the total It-electron energy is 6a + 6E3 (M.O.) and Q + 2.61 J (V.B.). The total %-electron energy of the hypothetical cyclo- hexatriene is 6a + 8f3 (MX.) and Q + 1.5 J (V.B.). The resonance energy of benzene is, therefore:- R.E. = (6m + 6p) — (6a. + 813) . —2p (M.O.) R.E. = (Q + 2.61 J) (Q + 1.5 J) = 1.11 J (V.B.) J is a negative quantity, so that both methods indicate that benzene is more stable than the hypothetical cyclo- hexatriene. By inserting the value of 36 kcal./Mole.for the resonance energy of benzene, 0 and J may be determined. -11-

2p 36 kcal./mole. = 18 kcal./MOle. 1.11J = 36 kcal./mole. J = 32.4 kcal./Mole. The calCulation of resonance anergie6 in terms of p and J may be carried out for other aromatic molecules. Table 1(7)

J.Ip p keal./mOle)

BenZene 1.80 18 Naphthalene 1.80 16.6 Anthracene 1.80 16.2 Phenanthrene 1.80 18.2 Biphenyl 1.89 17.1

The constancy of the ratio JO indicates good agree- ment between the two theoretical descriptions of benzenoid aromatic molecules. The actual calculation of resonance energies fram the two theoretical descriptions is only empirical, as the value for p and J must be fixed from the experimental results for one molecule. Usually this molecule is benzene. -12 -

b) Experimental determination. The resonance energy of a molecule may be determined experimentally by comparing the calorimetrically deter- mined heat of combustion with a value derived from the summation of the bond heat increments for all the bonds in one w-electron localised structure of the molecule. For benzene, the resonance energy is obtained by subtracting the experimental heat of combustion (789.1 kcal./mole) from the summation of the increments for three a = C, three C C and six C - H bonds. Taking the bond heat increments given by Klarges (8), it follows that the calculated heat of combustion is:- (3_x 117.4) + (3 x 49.3) + (6 x 54) + 1 = 825.1 kcalide. (The added 1 kcal. is a correction for a six membered ring). This leads to a resonance energy for benzene of 825.1 - 789.1 = 36 kcal./mole. Because the resonance energy obtained in this way is a difference between large experimental values, it is subject to comparatively large error. For instance, an error of .1% in the determination of the heat of combustion of benzene would produce an error of 2% in the calculated resonance energy. A further difficulty is that the value derived from bond heat increments may be obtained from two different - 13 -

sets of values. One set is obtained by a straight- forward average for the bond type taken from the experi- mental heats of combustion of many non-aromatic compounds (9). The second set contains these same average values but with corrections made for the molecular environment of the bond (8). Whichever set is used makes little difference to the estimated resonance energy of benzene, but considerable discrepancies occur for polynuclear benzenoid hydrocarbons. A similar procedure allows the calculation of resonance energies from the heats of hydrogenation. Heats of hydrogenation have been determined for only a few aromatic compounds. This is unfortunate as they give the more accurate results. It is not surprising, therefore, that resonance energies are usually quoted as falling within a range of values (Table 2). - 14 -

Table 2

Resonance Energy Resonance Energy kcal./mole. kcal./k-electran

Benzene 36 6 Naphthalene 61-75 6.1 - 7.5 Anthracene 86-105 6.1 - 7.5 Pyrrole 21-24 3.5 - 4.0 Furan 17-21 2.8 - 3.5 o Thiphene 29-31 4.8 - 5.2 I. Pyridine 23-42 3.8 - 7.0

From Table 2 it is apparent that there is little correlation between chemical reactivity and resonance energy per %- electron. The polynuclear benzenoid hydrocarbons have a higher resonance energy per %-electron than benzene but they are much more reactive than benzene chemically. On the other hand, thiophene hasa smaller resonance energy than benzene but is more reactive than benzene. However, compounds which do have a large resonance energy show at least some of the chemical properties of benzene. A small resonance energy and typically olefinic chemical properties, together would inevitably classify a compound -15- as non-aromatic. The use of resonance energy as a criterion of aroma- ticity may have its limitations but this criterion has proved consistently more successful than any other in detecting and estimating aromaticity. There is, for instance, no consistent way of using carbon to carbon bond lengths in a molecule as a criterion for aromaticity (ii, 12, 10). A polynuclear benzenoid hydrocarbon may have some carbon to carbon bond lengths which approximate in length to those in aliphatic hydro- carbons but may still possess a large resonance energy. Indeed a polynuclear compound may show a wide variation in carbon to carbon bond lengths around its moleaule. Again, the benzene molecule is planar to within a few hundredths of an angstrom unit whereas in the molecule of [18] annulene the six inner hydrogen atoms are dis- placed by molecular overcrowding by 0.9k perpendicular to the mean plane of the molecule (13). Yet [18] annulene is considered to be aromatic (14). Apart from the evidence of its n.m.r. spectrum, it is considerably more stable than [16] annulene or [20] annulene. A molecule in whidh the distortion from planarity is much larger, would not be expected to have a large resonance energy, for no overlap of the 2 pz orbitals in the molecule could -16-

ocaur and, therefore, no delocalisation of %-electrons would be possible. Cyclooctatetraene is such a molecule and the molecule has a tub-shaped structure. It is possible to classify aromatic compounds according to theoretically derived criteria. The Mickel rule (15), which applies only to monocyclic hydrocarbons where the it-electron centres are arranged on a circle, classifies as aromatic those molecules having 4N + 2 7c-electrons where N is 0,1,2,3 The rule, derived from early N.O. theory, is firmly based theoretically and has been experimentally verified for a number of compounds (16,17). There is no basis for an equivalent rule in the V.B. description of aromatic compounds. Craig's rule (18) is based on the symmetrical nature of the V.B. ground state wave function of all known aromatic compounds. To apply Craig's rule, a structure of the molecule under consideration is drawn with localised 7-election centres. This structure is then labelled with equal nuMbers of the spin symbols, a and p; at each end of each double bond, placing them alternately as far as possible. For benzene the labelling is thus:- - 17 -

The whole Molecule is now rotated, abaUt a two-fold axis of the mOlecUle, through t radians. The tWo-fOld axis Chosen must pass through at least two of the 7C-electron centres. The charaCter of this transforMation On is giVen by (-1)P 4 where p is the number of interchanges of 7c-electron centres effected by the transformation and q is the number of interchanges of spin symbols required to restore the original labelling system from the transformed one. Aromatic molecules transform with a positive sign, whereas for non-aromatic molecules the character of the transfor- mation is negative. Por benzene, p = 2 and q = 0, hence:- (..1)2+o

= +1 . AnY classification of aromatic moleaules which is based on the properties of the molecular ground state of the mole- cule, as is classification based on resonance energy or either of the two theoretical rules given above, will not be expected to correlate with a classification based on chemical reactivity. The chemical properties of art aromatic compound depend very largely on the properties of the transition state of the molecule during chemical -18- readtion, The properties of the molecular groUnd state have only a secondary influence on the chemiCal properties of an aromatic compound. - 19 -

CHAPTER 2

NUCLEAR MAGNETIC RESONANCE AND AROMATICITY

Introduction Nuclear magnetic resonance spectroscopy is the study of the absorption by nuclei of electromagnetic radiation in the radio frequency range. In this thesis, n.m.r. refers to proton magnetic resonance only. The general theory of n.m.r. has been expounded in several standard works (19, 20, 21) and it is proposed to discuss only the relationship between the behaviour of aromatic molecules in a strong magnetic field and the n.m.r. spectrum of such molecules. Diamagnetic Anisotropy and Molecular Susceptibility. Nearly all diamagnetic molecules are magnetically anisotropic. Such molecules, when placed in a non-uniform magnetic field, will move from a position of higher magnetic field to one of lower. A quantity, known as the bulk molecular susceptibility, may be derived by measuring the apparent mass change when a sample of com- pound is moved from an area of zero magnetic field into one of strong magnetic field. The strong magnetic field - 20 - must be non-uniform. If the sample were moved into a uniform magnetic field, no apparent mass change would be observed. From such measurements on a very large number of compounds, Pascal (22) concluded that the molecular suceptibility (xin) of a molecule was represented by the expression:-

where na is the number of atoms of susceptibility, y'a, in the molecule and X is a correction factor representing the nature of the bonds between the atoms in the molecule.

The constant Xa and the correction factor X were derived empirically from measured molecular susceptibilities (Table 3). Pascal's Constants for Atoms - Table 3.

Atom & x 106

H ... 2.93 C - 6.00 N (open chain) - 5.55 N (ring) - 4.61 - 21-

Corrections for Bond Types.

Bond T. x 106

C = C + 5.5 C=0.,..0=0 + 10.6 ,-CmN + 0.8

Considerable success in predicting molecular susceptibilities has been achieved by the use of Pascal's constants. The Pascal sum for the hypothetical cyclohexatriene gives the molecular susceptibility for the molecule of -36.9 x 166 e.m.u./mole. The experimental molecular susceptibility of benzene is -55.6 x 106 e.m.u./mole. The difference between these values is -18.7 x 166 e.m.u./mole and it is known as the exaltation. The results of similar calculations for other molecules are shown in Table 4. Table 4 (7)

Molecular Suscepti- Exaltation bility x 106 e.m.u/ x 106 Compound mole. e.m.u./mole. Expt. Pascal Sum Benzene -55.6 -36.9 ' 18.7 Cyclooctatetraene -51.9 -49.2 2.7 Naphthalene -91.2 -55.7 36.2 Diphenyl -102.9 -68.0 34.9 - 22-

Prom Table 4 it is seen that exaltation provides another criterion of aromaticity in much the same way as resonance energy. The exaltation for cyclooctatetraene is little more than that expected for four simple conjugated double bonds, and on this basis cyclooctatetraene would be expected to be non-aromatic. The magnetic susceptibility of a molecule, as normally measured, is the average of three principal susceptibilities directed along three mutually perpendicular axes. For benzene, the principal susceptibility along the axis perpendicular to the plane of the ring is approxi- mately twice that along the other two axes. Pauling (23) explained this phenomenon. The it-electrons of benzene were considered to be able to circulate freely around two ring-shaped regions above and below the plane of the aromatic molecule. The influence of a magnetic field, applied in a plane perpendicular to the plane of the aromatic molecule, causes these electrons to circulate. Pauling estimated the contribution of each %-electron to the total suceptibility of the molecule and a summation of these contributions gave the required, large suscepti- bility along the axis perpendicular to the plane of the aromatic molecule . -23-

Pauling's model of electrons flowing in a closed circuit is only a useful physical approximation. The true effect of the applied magnetic field is the induction in each electron of a precessional motion. The combined effect of these motions for all the It-electrons is as if they circulate around the c-bond framework of the molecule. It should be pointed out that in the absence of an applied magnetic field the it-electrons' motions are effectively random within their orbitals. Nuclear Magnetic Resonance. The chemical shift of benzene protons is 2.731:. These protons resonate at considerably lower applied field strengths than olefinic protons in an analogous environment, which have a chemical shift of 4.267 (cyclooctatriene). Similarly a difference in chemical shift is apparent between the methyl protons of toluene (7.66 -r) and those of an in-chain methyl group in a carotenoid (8.061:). Pople (24) used the simple Pauling model (Fig.1) Fig. 1

• •

• • /1;11/ - 24-

to explain these anomalous chemical shifts. At the periphery of the aromatic ring, the field (H1), induced by the circulation of the ,t-electrons, reinforces the applied field (H0). The protons attached directly to the ring or the protons of a substituent methyl group, would thus come to resonance at a lower applied field than would otherwise be the case. The hypothetical circulation of n-electrons became termed "a ring currentic Pople equated the magnetic effect of the ring current to an equivalent point dipole placed at the centre of the aromatic ring. This allowed the calculation of the induced field at points external to the aromatic ring so that the chemical shifts of attached protons could be derived. The point dipole approximation could be improved by considering that the it-electrons circulated in two rings approximating to two at-orbitals, one above and one below the plane of the aromatic molecule, as was considered by Fessenden and Waugh (25). The theory advanced by these authors predicted that the protons held over the centre of a benzenoid ring should be shielded rather than deshielded. They examined the spectra of [12] paracyclophane and [10] paracyclophane and found that the methylene protons, in the middle of the bridging chain of methylene groups and - 25 - held over the centre of the benzene ring, were indeed shielded, as the theory predicted. The ring current theory had thus led to verifiable predictions and so became firmly established. Minor errors in the theoretical calculations of Fessenden and Waugh were corrected by Bovey and Johnson (26) and these latter workers calculated the shielding in three dimensions around the benzene molecule. In addition, Bovey and Johnson noted that the diamagnetic shielding region was larger than the paramagnetic. This correlated with the upfield shift of protons in any compound when the solvent in which the measurement of the chemical shifts were made was changed from a non-aromatic solvent to an aromatic one. The above form of the ring current theory has been improved by Pople (27) and by McWeeny (28). Benzene is a compound which is firmly established as aromatic by any criterion, and its n.m.r. spectrum has been satisfactorily interpreted in terms of a ring current theory. Now the process may be reversed and an aromatic compound could be defined as one whose molecule is capable of sustaining an induced ring aurrent. This was the approach of Elvidge and Jackman (29). This definition has proved very useful. The work of Sondheimer and his co-workers (14) is particularly - 26 -

instructive. The n.m.r. spectrum Of [18] annulene

shows 12 protons at a chemical shift of 1.1 and six protons at 11.8 The six protons are the "*icier" protons and these are held in the diamagnetic Shielding region. The 12 protons are the "outer" protans and are held in the paramagnetic shielding region. Thus [18] annulene is considered to be aromatic as it is capable of sustaining an induced ring current in its molecule. Estimation of Aromaticity. In addition to the definition of aromaticity des- cribed above, Elvidge and Jackman (loc.cit.) suggested that the magnitude of the induced ring current in a par- ticular molecule might be a measure of the degree of aromaticity of that molecule. The magnitude of the ring current in the molecule in question could be compared with that in the benzene molecule, and a fractional aromaticity for the molecule derived relative to benzene. -27-

To estimate the magnitude of the ring current in benzene, the chemical shift of the methyl protons in toluene (7.661) are compared with that for an in-chain methyl group in a carotenoid (8.06T ). This latter methyl group is in an approximately similar environment to the former, except that no ring current is possible in a carotenoid. The methyl substituted carotenoid is known in this con- text as a non-aromatic model for benzene. The ring current contribution to the shielding of the methyl group in toluene is 0.4 . The chemical shifts of methyl groups in other aromatic molecules are easily determined, but non-aromatic models for these aromatic molecules are not always available. When this occurs, estimated values for the chemical shift of the methyl groups in the absence of a ring current must be deduced from, available data. Elvidge and Jackman (loc.cit.), in their estimation of the aromaticity of 2-pyridone, used the shifts of methyl groups attached to this ring. Methyl groups were preferred as their chemical shifts are not so sensitive to small changes in local environment as those of protons attached directly to the aromatic ring. The non-aromatic models (11,111) were:- - 28 -

An alternative approach used by these workers was to subtract the ring current contribution from the shifts of methyl groups attached to pyridine and to pyridinium salts. The ring current contribution was assumed to be the same as in benzene. Comparison of these calculated chemical shifts with experimentally determined ones for the methyl groups attached to 2-pyridone gave a second value for the fractianal aromaticity of 2-pyridone. These two approaches gave a fractional aromaticity of .36 and .31 respectively (Benzene = 1) for 2-pyridone. When the first of these two approaches is applied to the estimation of aromaticity in other aromatic molecules, the principle difficulty lies in the choosing of suitable non-aromatic models. Recently, two attempts to estimate the aromaticity of thiophene and furan have appeared in the literature. Elvidge (30), and Abraham and his co- workers (31) use the same equation to calculate fractional aromaticity. For furan the expression is thus:- - 29-

rind; current in furan fractional aromaticity - ring current in benzene

AB(177F)3 AF RB where AB and bp are the differences between the experi- mental chemical shift of a proton in a benzene or furan system and the estimated chemical shift of a similarly located proton in a non-aromatic model of each system;

AB and AF are the areas of the benzene and furan molecules respectively; RB and Rr are the distances of the proton from the centre of the benzene and furan ring respectively. The protons may be attached directly to the aromatic molecule or may be contained in a methyl group similarly attached. The latter are preferred. The difference between the results of Elvidge, which correlate with the resonance energies of the molecules concerned, and those of Abraham, which do not, stems from the choice of their respective non-aromatic models. However, Elvidge was able to extend the scope of his calculations to pyrrole (Table 5). Table 5

Fractional Aromaticity Compound R.E.—57- Elvidge Abraham Thiophene .81 .75 .90 , . Furan .45 , .46 .94 Pyrrn1R _An _qg — — 30 —

An alternative ap.roach. The methods described above for the estimation of fractional aromaticity of aromatic molecules depend on measuring the chemical shift of protons held in the para- magnetic shielding region. A mid-chain methylene group in [10] paracyclophane is held in the diamagnetic shielding region. If the benzene ring in [10] paracyclophane is replaced by another aromatic ring, the methylene group will still be shielded but the degree of shielding will depend only on the relative magnitude of the induced field in the new aromatic ring. Allowance must be made for the fact that the shielding effect varies with the position of the methylene group relative to the aromatic ring. ”Iso- dhieldingu diagrams, such as that of Bovey and Johnson (loc.cit.), should prove useful in this respect (Fig.2). Fig.2 (26)

p reNdiu.$) - 31 -

The diagram represents one quadrant of a plane passing normally thraugh the centre of the ring. The iso- Shielding lines represented the shift in n.m.r. shielding value which will be experienced by protons as a result of the induced field of the benzene ring. The methylene group held over the aromatic ring is in fact being used as a probe to estimate the induced field at, a partiaular position in space around the aromatic ring. The methylene group used as a probe must be separated from the aromatic ring by a sufficient number of methylene groups to ensure that it is free of any shielding effect of the aromatic ring other than that due to the induced field of the ring. Fortunately shielding effects arising from inductive and mesomeric effects and from charge density are not trans- mitted along chains of saturated carbon atoms for further than two or three atoms. The spatial arrangement of the methylene chain relative to the aromatic ring may be determined by consideration, of accurate atomic models of the molecule concerned. - 32 -

CHAPTER 3

THE SYNTHESIS

Introduction. The paracyclophanes have two analogues in the field of bridged pyridines. These two analogues may be represented by the structures (IV and V).

x

Of these, the free base (IV) is the more desirable as, by Tuaternising the pyridine ring with an alkyl halide, a system analogous to that of the Quaternary compound (V) may also be obtained. In this work a twelve-membered methylene bridge (m = n = 12) was chosen in both attempted syntheses. This would allow easier comparison with the work of Waugh and Fessenden (25) as one of the compounds whose n.m.r spectrum was studied by these authors was [123P aracyclo- phane. The twelve-membered bridge was chosen in preference to the ten-membered as it was felt that, in view of the -33-

methods envisaged for cyclisation, the larger ring represented the better chance of success. In the attempts described in this thesis to synthe- sise the bicyclic system (IV), the experimental work has been concentrated upon the preparation of a dialkyl pyridine of the type (IVa).

C N 2) • t vr-1

where X and Y represent two functional groups which would permit ring closure. The values of n and m must be adjusted to give a twelve-membered methylene bridge in the final bicyclic product. No attempt has been made to prepare substituted derivatives of either compounds (IV or V). It is pertinent to review methods for the synthesis of macrocyclic compounds and to examine their relevance to the synthesis of bridged pyridines. - 34 -

1112111.22212E112eSthes,crocr odsun. a) Acyloin Condensation (32). The acyloin condensation has proved to be a most effective and versatile method for the synthesis of macrocyclic compounds. It involves the intramolecular condensation of an cy1)-dicarboxylic ester and the product is a cyclic acyloin.

NA,

The yield of acyloin is good and even when n = 7-11 it does not fall below 40%. All other methods for the synthesis of macrocyclic compounds give very low yields when they are used to prepare macrocyclic compounds with 9-13 members in the ring itself. The acyloin condensation was considered to be inapplicable where a pyridine ring was present in the diester molecule. Under the influence of alkali metals, pyridine itself undergoes radical coupling reactions of the type:- -35-

iN1 0 o,

tV t\VO. These radical coupling reactions proceed rapidly and the most likely result of an attempt to cyclise a pyridine diester of the type (IVa; X = Y = CO2Et) by the acyloin condensation would be the formation of a series of dipyridyls and polymeric products. The acyloin condensation was used by Cram and Daeniker to prepare paracyclophanes (33). One striking feature of this work was the failure of the acyloin condensation to produce [14]paracyclophane, although no difficulty was encountered in the preparation of [123para- cyclophane. Other bridged aromatic compounds have been prepared by the acyloin condensation and these include those for thiophene (34) and biphenyl (35). It appears quite possible to use the acyloin conden- sation in the synthesis of bridged pyridines by forming the pyridine ring subsequent to the formation of the macro ring. This is discussed in Chapter 4. Ruzicka's Method (36). Ruzicka obtained macrocyclic ketones from the -36-

pyrolysis of yttrium or cerium salts of a,(,)-dicarboxylic acids. The yield of ketone was very small (< 2%) and for this reason alone the method was rejected as a possible route to bridged pyridines. Tas....11101-"...12L-RS1242z1-1191L-ILS2a1U- In the synthesis of macrocyclic compounds from acyclic a+w-difunctional compounds there are, inevitably, two competing reactions. One is the intramolecular reaction which gives the desired product and the other is the intermolecular reaction which gives polymeric product. To suppress the latter reaction, the concen- tration of the difunctional compound (c) in the reaction media must be as low as possible. The kinetic explanation of this principle has been discussed by Bennett (37). Essentially, the intramolecular reaction has first order kinetics. Thus:- - de _ kfc aT - and the half-life (t') of this type of reaction is:-

= 27- ln 2

The intermolecular reaction has second order kinetics. Thus:-

= kolc2 and the half-life (t") of this type of reaction is -37—

too = 1

If TTt" is. large, a good yield of the macrocyclic compound may be expected. Hence:- t" k° 1n2 1-r = 271 —

= 1n2K-1 . Thus, reaction mixtures in which the concentration of difunctional compound is small will favour the intra- molecular reaction. Obviously high dilution may be obtained by using a large volume of solvent in the reaction mixture. If the condensation of the difunctional compound is irrever- sible, the same dilution will be achieved by adding the difunctional compound very slowly to the reaction mixture (which does not require to have a large volume). Theoretically any bond-forming reaction may be adapted to the synthesis of macrocyclic compounds. In practice, some reactions, which readily yield five or six membered cyclic compounds,often fail to yield macro- cyclic compounds (but cf. ref. (39)) (e.g. the aldol condensation). Other reactions apparently cannot be adapted to yield cyclic compounds of any ring size (e.g. Kolbe electrolytic synthesis (40)). - 38 -

Ziegler was the first to make use of the high dilution principle, which is employed in all the methods described below. Hunsdiecker Method (41). This method involves the intramolecular condensation of an Gl-bromoacylacetic ester, and has been used in the prsparation of bridged thiophenes (42)

C 0 H 2)r, k ©AI./ C.H.CO,D-I CH,

The method was not thought suitable for the synthesis of a bridged pyridine as a campeting reaction would be the quaternisation of the pyridine by the alkyl bromide. The pyridine could be quaternised with a separate bromide before cyclisation but the quaternising group would have to be removed subsequently. Actually, later work showed that in dilute solution alkyl bromides quaternise pyridines reluctantly. -39-

Friedel-Crafts/ Reaction. A high dilution adaptation of the Friedel-Crafts/ reaction has been used to prepare paracyclophanes (43).

MC13 (c11 11-01,

COCA \\,-5(7) This method, in its above form, is not applicable to the preparation of bridged pyridines as pyridines do not undergo a Friedel-Crafts/ reaction in the accepted sense where the aromatic molecule is the nucleophile. In Chapter 4, a method for using the Friedel Crafts/ reaction in the required synthesis is suggested. Diketene method (44). The diketene method has no features which would disqualify it for use in bridged pyridine synthesis.

EE (CH 2)t) (C Hz) • n•-•-2, COCk L--- CH CC C)

/n I I _ — o

- 40 -

The preparation and handling of pyridine acid chlorides requires care. The yields, however, are inferior to those of both the Hunsdiecker and the Ziegler method and the diketene method gives polymeric products very easily. Ziegler method 5i5) The possibility of using the Ziegler method has been investigated experimentally in this thesis.

C (et,. CH.CN Ev. P111. ) , C C N

PrOvided a suitable dinitrile (VI) could be prepared, Ziegler's method was thought to provide an acceptable route to the bridged pyridine (IV).

(cH0.CINI 2)„L, = I V 41 -

In practice,the synthesis of the pyridine (VI) was not as straightforward as had been hoped. An attempt to repeat some of Ziegler's work suggested that, for the successful intramolecular cyclisation of the dinitrile (VI), his method would have to be modified. This feature is di s- cussed later. Diacetylcue Method (46 14) . A more recent method than those described above for the preparation of macrocyclic compounds involvesthe intramolecular oxidative coupling of the a,,€-diacetylenes.

C;CH

Ckk (0 (b-c) C C C C

The method has been used extensively by Sohdheimer and his co-workers (14) in their studies of poly unsaturated madrocyclic compounds. [18]AnnUlene has been mentioned several times in this thesis. It was obtained by Sondheimer Wolovsky and Ainiel (47) by the following synthesis:

- 42 -

C-401ke...) p ;AS

KO Flu('

PdC

The diacetylene method is applicable to the synthesis of saturated macrocyclic compounds by the complete reduction of the 1.13-diyne which is formed in the con- densation. Indeed, patents have been taken out for the use of this method in the manufacture of macrocyclic perfumery materials (48,49). The synthesis of a suitable pyridine diyne (VII) was much simpler than that of a pyridine dinitrile (VI).

ck-1 HC 16(c-1-12) VII

- 43 -

It could not be predicted from experiments in the literature whether or not the diyne (VII) would couple intramolecularly to give the monomer (Throughout this thesis, the terms Ilmonomer", lEimerle v eto., are used as convenient approximations to denote the order of molecular complexity of the cyclic product derived from the true acyclic monomer). The diyne (VIII) coupled to give the dimer and, possibly, a little monomer (46). The diyne (IX) produced only dimer and higher polymers (50).

C.Vi-CiCHJ.: 2. •

CVA .C\-\2 C :CH Jx An accurate atomic model of the molecule of the conjugated diyne (VIIa), whiCh would be the initial product of the intramoledular coupling of the diyne (VII), showed that such a molecule was Completely strain-free.

VI i ck. - 44 -

The pyridine diyne (VII) would be fairly soluble in slightly acidic, aqueous alcohol owing to the solvating effect of the quaternised nitrogen. Although most of the coupling reactions, in which cyclic products have been formed, have been run in pyridine solution, aqueous alcoholic solutions had been widely used for acyclic coupling (51). Use of aqueous alcoholic solutions rather than pyridine for the coupling of the diyne (VII) allows an alternative approach to the cyclisation. Other Methods. There are several methods for the preparation of macrocyclic compounds in which a hetero atom is incor- porated into the ring (52, 53). The preparation of macrocyclic amides (52) is a good example of the application of high dilution techniques to a reaction in which the products are usually acyclic. Methods for the Synthesis of. the Puidinium Salt (V). The foregoing survey of methods for the synthesis of macrocyclic compounds has been restricted to the 2,5 bridged pyridine system (IV). However, the methods should apply equally to the synthesis of the 1,4 bridged system (V). For instance, instead of the dinitrile (V4 a quaternary dinitrile (X) could be employed in the application of Ziegler's method. -45—

C . c. N rt,cn

h rrt + Br: LC .-.1z.•( ‘7.0%;,C_ N X

In the present work, all the attempted syntheses of compounds of the type 00 were from compounds of the type XI, which it was hoped would undergo intramolecular quaternisation.

CH14. .7) . C. Fizx 10 on"

••••••.•

A

; n 46 -

The 20 Bridged Pyridin Scope of work. As pointed out in the preceding discussion, aldol type condensations are not generally suitable for the preparation of macrocyclic compounds. At the beginning of this work, it seemed a pity to neglect the protropic activity of a 2-methyl group of a substituted pyridine. Such methyl groups condense with aldehydes and it was hoped that the pyridine aldehyde (XII) would undergo an intramoleaular condensation.

C HoN (C ,CHO

Me XI i

However, whereas aromatic aldehydes condense readily with 2-picolines, aliphatic aldehydes do so reluctantly (54). The condensation might be assisted by prior quaternisation of the pyridine ring but there are two objections to this idea. Firstly, the quaternising group must be removed after condensation and secondly, any group attached to the pyridine nitrogen is likely to produce some steric hindrance towards the approaching formyl group.

- 47 -

Several months work was carried out on the above approach until it was abandoned in favour of one based on the Ziegler method for the preparation of macrocyclic compounds. Later, work was commenced on an approach based on the synthesis of macrocyclic compounds by the oxidative coupling of terminal diacetylenes. In fact, until the diacetylene approach began to yield encouraging results, work on both approaches continued concurrently. Attempted synthesis of 5-101-formyldecy1-2-methylpyridine. The first synthetic route investigated was as below:-

OTs

XIII XIV

:(C.1-k2). CNz, Ni3C

cr‘, ckr\,) c HO ict X V

The tosylate (XIII) appeared to be aCcessible by the route shown below:-

CO ) Me't, XVI XVII. -4'g—

cox Et XVIII 1L,A, V4 cl-t,O \-1

Although the alcohol (XIX) was prepared without difficultYr the tosylate was not isolated by treatment of the alcohol (XIX) with p-toluene sulphonyl chloride under a variety of conditions. Although benzyl tosylates are only preparable under special conditions (55), there seemed no reason to suspect that tosylation was not taking place, especially as m-nitrobenzyl tosylate has been prepared (56). The difficulty probably lies in the work-up of the reaction mixture. The only compound isolated from the reaction was the recovered alcohol (XIX). Fortunately, a new approach to the preparation of 5-alkyl-2-methyl pyridines was found. Plattner and his coworkers (57) had been able to prepare these pyridines from 6-methylnicotinonitrile (XX) and suitable Grignard reagents. In this way

- 49 -

5-dodec-11t-enyl-2-methylpyridine (XXII) was readily prepared.

)C-1 Me4 V- \ 4_ XX C C (C_qiC141 Ply;

The yields on the two stages were 56% and 73% respectively. Generally, on the ccendensation stage, slightly better yields than Plattner's (46% loc.cit.) were obtained by a slight modification of his method. There was no necessity to purify the ketone (XXI) before reduction and the reduction always eliminated traces of the starting nitrile (XX) which were difficult to remove otherwise. Ozonolysis of the olefin (XXII) might have yielded the aldehyde (XII) but there is a danger of completely cleaving the pyridine ring (58). The olefin (XXII) gave a low yield of the dihydroxy derivative (XXIII) by the Pr6vost reaction.

CA-l. 0-12.011 71. 63nze. 116. XXI -50 -

By the time this stage had been reached, serious doubts had arisen as to whether the aldehyde (XII) would cyclise satisfactorily. However, the diol (XXIII) was treated with periodate and the product was shown to be an aldehyde, from its infra-red spectrum

(0max 865 1708 cm.-1). Although this approach was abandoned at this point the experience gained was applied to later work successfully. The availability of the olefin (XXII) was to prove very useful when a model com- pound was required in the diacetylene synthesis. Ziegler's Method. Application of Ziegler's method required a dinitrile (VI). The synthetic problem was approached in two ways. Firstly (Route 1), the dinitrile could be prepared in the last stage of the synthesis.

OZ-A- ) VI 13ir XXI X X V

Secondly (Route 2), the two nitrite groups could be incorporated at different stages of the synthesis.

- 51 -

NI (c H2) .CN It XX V I X X \n / N

Cry. H XXV1!(

X is some function easily replaced by cyanide. The preparation of the pyridine (XXVI) is from acyclic materials. Route 1. This approach requires a 2,5-dialkyl pyridine as opposed to a 5-alky1-2-methylpyridine as before. Alkyl- ation at the 5-position had already been investigated. Alkylation at the 2-methyl can be carried out by activation of the methyl group by phenyl lithium followed by condensation with an alkyl halide (59). The choice of ethoxy halides was governed, in the first instance, by availability and 2-chloroethyl ethyl ether (XXIX) is the most readily available. As p-chIoro ethers do not form Grignard reagents (60), the ether (XXIX) would serve to aakylate the 2-methyl group. Thus a 7-halo-l-ethoxy heptane was required for the 5-position. 7-Bromo-l-ethoxy heptane was duly prepared and condensed, as its Grignard reagent, with 6-methylnicotinonitrile ( .

-52

(-1.); t x>( , ‘vtlbt- XXX XXX I

The preparation of a model for the next stage seemed desirable. 5-Ethyl-2-methyl pyridine (XVI) was readily available and this was condensed, using phenyl lithium, with 2-chloroethyl ethyl ether (XXIX). The ethereal linkage of the resulting ether (XXXII) was cleaved with hydrobromic acid to give the hydrobromide (XXXIII) but the nitrile (XXXIV) was not isolated although a variety of reagents and conditions were investigated.

1)XXIX X\n

\ Br

Br XXXttl -53-

The product from the attempted preparation of the nitrile was always soluble in aqueous alkali. This suggests that an alkyl pyridinium salt had been formed. The hydrobromide (XXXIII), in fact, could be imagined to form an intramolecular quaternary salt (XXXV) very easily.

( X X-rn In the alkylation at the 2-methyl group, the method above used a ratio for Phld:base:halide of 2:2:1. This produced a 73% yield of alkylated product (XXXII) based on halide. This was not very economical when the base is precious and the halide readily available. So a ratio PhLi:base: halideof 1:1:1 was used when 2-methy1-5.980-ethoxy octyl pyridine was alkylated. This gave a yield of dialkylated pyridine (XXXVI) of 30%.

DE-• XXXI. _22 Pv1--k xx IX FtO c •••••••••..--..irm......

XX XV -54-

Unfortunately very little starting material (XXXI) was recoverable. This difficulty at a vital stage in the synthesis of 2,5-dialkyl pyridines was later solved during the synthesis of 2,5-dihex-51-enyl pyridine. To avoid reactions of the type XXXIII XXXV a different chain length was required. By an exactly analogous route 2-61-ethoxyhexy1-5-51-ethoxy pentyl pyridine (XXXIX) and a model compound 2-6F-ethoxyhexyl- 5-ethyl pyridine (XL) were prepared.

>TX co, (c.‘42.)4:

EI-70 . 11%r. xxx t

Eta XX X 1)( X VIII -55-

This completed the work in this field as by this time the diacetylene approach was yielding encouraging results. Essentially the wide applicability of this method of synthesising 2,5 dialkyl pyridines had been demonstrated. Later work suggested several improvements. Firstly, alkylation at the 2-methyl group is best performed using potassamide in liquid ammonia. Secondly, it appears that the literature methods for the preparation of pyridine alkyl nitric es from the corresponding bromides are not the best (61). A better method may be to treat the chloride with sodium cyanide in dimethyl sulphoxide (bromides react with the solvent). To prepare a suitable chloride it might be necessary to proceed via the alcohol.

a X L SOClz conc., Ct

Again, a more readily cleaved ether such as a Biphenyl methyl ether might be used.

— 56 —

Route 2. This approach to the synthesis of a suitable dini- trile involved the actual synthesis of the pyridine ring. „0,4-

X. CI-V,(C C C3. C No CH= OA H.C.01,E1- X , C (C.‘-1))il XL_I —)(71- claim otatamide

RCN

X.Cli z. (0-0,14 0 Where n = 10 or 9 According to Mariella (62) formylation of an unsymm- etrical methyl ketone with sodium and ethyl formate takes place on the methyl group rather than the methylene. The simplest way to incorporate X was to use a terminal clefinic ketone (XLIV) 0 CH2 = cHscH2.0H0 . C — CH3 " n-2 XtIV

The other suitable way was to use an ethereal group but this had to be cleaved under conditions which would not affect the cyano group already on the ring. Triphenyl -57— methyl ethers are cleaved to give halides on treatment with the appropriate phosphorous halide (63). This method would have the added advantage that the 2-pyridone (=II) would become a 2-halo pyridine concurrently. Accordingly, two unsaturated methyl ketones and one hydroxy methyl ketone were prepared. These were dodec- 11-en-2-one (XLIV; n = 9), tridec-12--en-2-one (XLIV; n = 10) and 13-hydroxytridec-2-one (XLV). The starting material in the synthesis of each of these three methyl ketones was undec-10-enoic acid. The synthetic routes were as below:- SO C1 CH2:CH4CH218CO2H CH2:011.(CH2).80001 Mel , LI CH2:013.(CH2)8.00.CH EtOH HBr 3 CH2:CH. 0112)8.0O2Et ;XLITI n = 9 BrCH COC1 Xr7T1 2' (CH2 )9. LII L.A.H. L.A.H. CH2:CH.(CH2)8.CH2OH BrCH OH ) .CH OH XLVIII 2' 2 9 2 LEI SO C1 2 KCN CH2:CH.(CH2)8.011201 C.N.CH CH ) CH OH IL 2e( —2 . 91 —2 - LIV KON MeMgBr -58-

0 0H2:CH.(0H2)8.CH2CN 0113.c.(0H2)9 .01120H IT 7M7 NeMgI 0 0H2:011.(0112) 9 .0.0113 2MTV. n = 10

Both of the unsaturated ketones (XLIV; n = 10, XLIV; n = 9) were treated with ethyl formate and sodium sand and the resulting sodium salts of the p-ketoaldehydes were treated with cyanoacetamide in aqueous ethanol. It was expected that formylation would take place on the methyl group of each ketone so that the pyridones, (LV) and (LVI),would be formed.

CN cr4 zt ). (3 C CH,: CH. (c H2) cli 8 H LV LV I The analytical results and the infra-red spectra for the products were in accord with expectation, but the n.m.r. spectrum of each product showed that a mixture of iso- meric pyridones had been formed in each case. Evidently,

-59-

some formylation had taken place on the methylene group as well as on the methyl group of each starting ketone. The n.m.r. spectrum of the mixture of the two isomeric pyridones derived from the ketone (XLIV; n = 9) is shown below.

9 b

jAtt\A kAjJ 3 5

e LV c L—c75 CH3 CM c ).0 H -&2: GI .H 2),c H CH,: CH. (0-12 6 7 H a

The pyridones (L,(LVII; n = 6) were present in the ratio (LIT4V11; n. = 0= 10:3. The n.m.r. spectrum of the mixture of the two isomeric pyridones derived from the ketone (XLIV; n = 10) showed similar characteristics and the ratio of the productsg.:94VII; n = )was 10:1. -60-

The desired pyridones (LV and LVI) are therefore the principal products of the reactions. Before attempting to separate the isomers in each mixture by chromatography, the mixture of pyridones (LV, LVII; n = 6) was treated with hydrogen bromide in the presence of added peroxide. Instead of the desired primary bromide, the secondary bromide was obtained. The n.m.r. spectrum of the crude hydrobromination product showed a sharp doublet at 8.321 from a terminal methyl group, CH3.CHBr.R., and a complex pattern at 5.95T from a methine proton, CH3.CHBr.R. The relative intensities of these two absorptions were 3 to 1. It might have been possible to circumvent these difficulties by using an W-bromoketone instead of an olefinic ketone in the condensation with cyanoacetamide. Treatment of the ketone (XLIV; n = 10) with hydrogen bromide in the presence of peroxide gave a mixture of primary and secondary bromides.

H 2 C:CH.(CH ) .__CO.__CH3 124 BrCH2.(CH2)10.00.CH3

+ CH3.011Br.(0H2)3.0. CO .CH3

Alternative methods for the preparation of 0 -bromoketones were not investigated. -61-

The alternative approach suggested at the beginning of this section was to use a triphenyl methyl ether of the type (LVIII) in the condensation with cyanoacetamide. Accordingly, 13-hydroxytridecan-2-one (XLV) was tritylated by the method of Pratt and Draper (64) and the oily product was shown by its n.m.r. spectrum to be the required ether (LVIII; n = 10). Ph3C.-0.CH2(0H2)n.00.CH3

LVIII This investigation was not taken further but the approach as a whole provides an alternative route, as shown below, to a suitable dinitrile for use in the Ziegler method of cyclisation.

1)formylation Ph3C0000H2O (0112)10 .00DH3 2)cyanoacetamide (LVIII; n = 10)

2) RCN C Br

-62-

Ziegler Method - Cyclopentadecanone. A snail part of Ziegler's work (45) was repeated in order to gain experience in the practical application of the high dilution principle. 1-Chloroundec-10-ene (IL) was available and this could readily be converted into 1,14-dicyanotetradecane (LXI). 1)Mg CH2:0H(CH2),C1 2) CH2:CHCH2Br CH2:011(CH2)10.CH:CH2

IL LIX HBr KCN ON.(CH2)14CN Br( )l er

LXI LX

1)Na+.N-B1.Ph (CH2)14 C = 0 2)hydrolysis

The principal experimental difficulty lay in designing an apparatus capable of adding a solution of the dinitrile to the reaction media over a period of several days. Ziegler's own apparatus was a complicated glass construction. The motive power was gravity. However, nowadays, elegant, electrically poweredI micro-addition apparatuses are available (65). These are rather expensive and a reason- able alternative is a graduated micro syringe with the ratchet removed and replaced by a cogwheel. Motive power -63-

is provided by a geared-dawn electric motor. The earlier apparatus, which was used in all the attempted cyclisations of the dinitrile (LXI), used a slowly rising mercury reservoir to displace a solution of the dinitrile by hydro- static pressure. This apparatus suffered from the effects of air-locks and of external temperature changes, and never operated smoothly over 3 or 4 days, as was required. The commercial apparatus has a wide range of continuous addition times and a large cubic capacity (125 mls.) and it would appear to be a sound investment where syntheses based on the high dilution principle are envisaged. Although a low yield of cyclopentadecanone was obtained using Ziegler's method, the problems of applying the method to a pyridine dinitrile grew larger as further experience was gained. The sodium (or lithium) salt of monoethylaniline is an unsuitable condensing agent when the dinitrile is a base. During the work-up of the reaction mixture, the monoethylaniline, which is present in a large excess over the dinitrile, may not be extracted with aqueous acid from the ethereal solution. The products Ziegler obtained were neutral ketones, and he was able to remove all the basic compounds from the reaction mixtures by extraction with aqueous phosphoric acid. Even so, Ziegler's isolation methods are involved.

- 64 -

It is felt that a "cleaner" condensing agent is required. The Ziegler method itself has not received much attention since its introduction, although the high dilution principle has been widely applied. A much wider variety of suitable condensing agents are available than was the case 30 years ago. Two suggestions are triphenyl methyl sodium and sodium hydride dispersion. Diacetylene Method. Some of the applications of this method to the syn- thesis of macrocyclic compounds have already been discussed. The first task was the synthesis of a terminal acetylene in the pyridine series which could serve as a model com- pound for the types of oxidative coupling reactions envisaged. 5-Dodec-ll'-enyl-2-methyl pyridine (XIV) was readily converted into 51-dodec-111-yny1-2-methyl pyridine (LXIII) H CH ) CHBr CH Br Br2 2.(--2•9° ---.--2-- XIV Me 001 LXIIT 4

H .(0112)9.CiCH NaNH2 liq.NH3 LXIII

The acetylene (LXIII) was then oxidatively coupled using a mixture of cuprous chloride and ammonium chloride in -65-

aqueous ethanol. Without isolationIthe intermediate l,3-diyne was hydrogenated to give (LXIV).

LXIII 1)CuOl/NH401/02> 4CHz)

2) H2/Pd Me Me

LXIV The yield on the two stages (LXIII LXIV) was 32%. This first result was distinctly encouraging. It was decided, therefore, to prepare 2,5 --cif.hex —51-3mylpyridine. 5-Hex-5'-eny1-2-methylpyridine (LXVI) was prepared by the familiar route:

CO.(0-t2). CH :CH2 3 _---4 CH:C11.(CH )•frtga LX V

Alkylation at the 2-methyl group of a 5-alkyl-2-methyl- pyridine using phenyl-lithium to activate the methyl group had been found to be unsatisfactory. Better results in such alkylation reactions are obtainable by the use of potassamide, in liquid ammonia, a reagent which has been used to activate the 3-methyl group of a pyridine -66-

(66) and which appears generally more effective than sodamide. Model alkylations of 5 ethyl -2 -methylpyridine (XVI) were carried out with the following results: 1)NaNH2/liq.NH3 XVI "h 2) BuBr 13,k. cH, 30% 1)10H2/liq.NH3> XVI Er 2) BuBr 7064, Bu • cH, Lxvli The contrast in yield was more striking when the two amide methods were applied to 5 -hex -51 -eny1-2 -methylpyridine (DWI): ain amide j Nt+s : Br CH; CH, (0-i2), LX V1

- NaNi-lx 6 t

K N Hi 4.77 0 Prom the dialkenyl pyridine (LXVIII) so produced, the required diyne (LXX) was prepared.

IA VI I 1 hickNH,. ) NH3

2),,cH5r.c1-tz8r BrCti .CH Br.(cHz)ii. -67-

The conditions which were used in the various attempts to couple oxidatively the diyne (LXX) are described in the experimental section. They cover most of the successful applications of oxidative coupling of diacetylenes in the synthesis of macrocyclic compounds which have been described in the literature. There was no evidence that any intramolecular coupling had taken place in any of the experiments. No coupling at all took place when a mixture of cuprous chloride and ammonium chloride in aqueous ethanol was used. In the other experiments, in which cupric acetate or cuprous chloride were used with pyridine as the solvent, only intermolecular coupling was observed. The mixtures of products, obtained from these latter experiments, were not fully analysed but theoretically should contain 2 cyclic dimers, 3 linear dimers, 2 cyclic trimers, etc. Clifford and Waters (67) suggested that the mechanism of the coupling reaction may be represented by the following reaction sequence. -68-

R.C: C 0 .7: 0 -/ Cu-)1I - OAo Me R.Ci C- CuII - 0- C= 0 + CuI0Ac Me R CfC' + Cu1OAc

R.Ci C.Ci C.R. If the above reaction sequence is a complete description of the coupling reaction, there is no reason to expect that the application of the high dilution principle to the coupling of the diyne (LXX) should fail to yield a cyclic monomer. However, Parkin and Coates (68) have shown that cuprous acetylides have a polymeric structure. For instance, X-ray crystallography gave the structure in Fig.3 as that of phenylethynyl(trimethyl phosphihe)CuI.

• C

\ • C

p PrMe 3

Fig.3. - 69 -

The trimethyl phosphine ligand would be replaced by pyridine in the coupling reaction. Atomic models were constructed for the diyne (LXX) and these indicated that if the two acetylenic groups of this diyne are brought into close proximity there would be little space above the pyridine ring in which to fit a cuprous complex of the type indicated above. If it is assumed that the radical, R.C;C'is never entirely free, and that it couples exclusively with a second acetylenic radical which was part of the same cuprous complex as the first radical, the failure to isolate cyclic monomer is explicable. The possibility of both terminal acetylenic groups of one molecule of the diyne (LXX) being present in a single cuprous complex is most unlikely on steric grounds. Thus no intramolecular coupling would be possible for the diyne (LXX). The above theory predicts that if the methylene chain of a diyne similar to LXX is long enough, cyclic monomers should be obtained by oxidative coupling.

-70-

Zreara....2.-422919212noundorStheticRoutes. a) Pyridine derivatives. A vital intermediate in much of the synthetic work in this thesis was ethyl 6-methyl nicotinate (XVIII), which was readily converted into 6-methylnicotinonitrile (XX).

c0,,E.t. C 0 Ni-kz kiu P003

XVIn L.XX

The ester itself was prepared by the oxidation of 5-ethyl- 2-methylpyridine (XVI), as described by Castle and Whittle (69). However, this method required the evapor- ation of 70 1. of water for the production of 200 g. of ester, and although the isolation of 6-methyl nicotinic acid, free from inorganic salts, was found to be entirely unnecessary, this extremely tedious procedure was frequently recommended in the literature (70,71). The early difficulties in the above route to ethyl 6-methylnicotinate (XVIII) prompted the investigation of an alternative route:- - 71 -

Na CH3.00CH3 + H* CO2Et CH3.00.CH2.CHO (as sodium salt)

Cif.CH2.00 .NH2

c POC-13 CI Pas H L.XXitt H 1.)01

However, the chloro pyridine (LXXIII) could not be selectively reduced to 6 -methylnicotinonitrile. An alternative approach, which was not investigated experi- mentally, was the following:

RwRiney CN Ni Vie

Yet another alternative, which was briefly investigated experimentally, involved leaving the 2 -chloro group intact until after the condensation of the nitrile group with a Grignard reagent. The pyridine (LXXIII) was used with butyl magnesium bromide and the product, which should be (LX7I7)/ was in fact shown to be a ketone by infra-red spectroscopy. -72-

co. Bu CN t BtA.M9By- me 1 ct z) aroltAsis Me Gt

LS. X ti L.XX0(

An interesting feature of this work was the isolation of a tetrachloride (LXXV) by treatment of the pyridone (LXXII) with phosphorous pentachloride. According to Mariella (72), the product should have been the mono- chloride (LXXIII).

CN PC15 _, c4\1 C. 'N 0 t-t Lxxik 1...Y X V

Laother approach to the synthesis of the ester (XVIII) involved the preparation of 5-acetyl-2-nethylpyridine (lama) NH Cl C O. C 4--\3 CH3.00.CH2.CH0 (as sodium salt) LXXV! -73-

The latter was then oxidised with concentrated nitric acid and the resulting acid esterified to give ethyl 6- methylnicotinate (XVIII). The overall yield was con- siderably lower than that from the oxidation of 5-ethyl- 2-methylpyridine. It had been hoped initially that the ketone (LXXVI) might have been oxidised directly to the 6-methylnicotinic acid by the Baeyer-Williger oxidation but this proved unsuccessful. The results obtained for a similar attempted oxidation on m-nitroacetophenone correlated with this observation (73). b) Aliphatic halides. A number of alkenyl and ethoxy alkyl halides were required and these were prepared by standard literature methods. Only a few features of these preparations are worthy of note. But-3-en-l-ol (LXXVII) was prepared by treating a mixture of magnesium and trioxymethylene with allyl bromide. The yield of alcohol was 30%. Linstead and Rydon (74) obtained a 45% yield of the alcohol by the same method but Ansell and Brown (75) were not able to obtain a yield greater than 10%. The latter workers introduced an alternative, three stage synthesis of but-3-en-l-ol but this was felt to be an unnecessary complication as the method of Linstead and Rydon is essentially simple. - 74 -

The usual literature methods (76,77) for the reduction of ethyl undec -10 -enoate(XLVII)to undec -10 -en - 1 -ol (XLVIII) employ the Bouveault -Blanc method. Much simpler and safer was the use of lithium aluminium hydride which afforded the alcohol (XLVIII) in very good yield. The addition of hydrogen bromide to ethoxy -olefins proved very effective for W-bromo-ethoxy-alkanes. The availability of ethoxy -olefins depends in turn on the availability of either a suitable alkenylhalide or a suitable ethoxyalkyl halide. 0H2:CH.(CH2)n.0112Br 2 5 CH0:0H.(CH ) "" 2 n+260X2H5° LXXVIII LXXIX

02115.0.01.12.(0H2)n.CH2Br 5.0.CH2( 0112 )1+2.CH:C},1

LXXX LXXXI For instance, 4 -bromo -1 -ethoxy butane (LXXX; n=2) is more readily available than 1 -bromohex -5 -ene so the second of the two methods was used to prepare 1 -ethoxyhept -6 -ene (LXXXI; n=2). l:Bz•ided tiumSalt. All the synthetic work in this section has been directed towards the synthesis of a compound of the type (XI). It was hoped, as pointed out in the introduction -75-

to this chapter, that this type of compound would undergo intramolecular quaternisation to give the bridged quaternary compound (V). The first attempt to prepare the halide (XI) used the following route:- Me ak2.(cH211.C.H:CH, 1)NaNH2/liq.NH3 2)1-chloroundec-10-ene LXXXI\

1\1 11-4 Br C C

J__XX II I t\) The secondary bromide (LXXXIII) was always the product of the addition of hydrogen bromide to the olefin (LXXXII) even when the addition was carried out in the presence of peroxide. The secondary bromide structure for the product(1,177III) was clearly indicated by the n.m.r. spectrum which showed a sharp doublet at 8.332 from the terminal methyl and a complex pattern at 6.0T from the .CHBr. proton. The secondary bromide (LXXXIII) could not be induced to quaternise intramolecularly. When it was heated alone under high vacuum, a diquaternary salt (LXXXIV) was isolated. - 76-

CH. CH..

CH3 . H, CHI

LXXXIV The main evidence for this structure was as follows: a) the molecular weight, determined ebullioscopically in benzene, was 669, whereas the calculated value was 653. It was assumed that the compound was not dissociated in benzene solution, b) the n.m.r. spectrum of the diquaternary salt (LXXXIV) showed no significant absorption above 8.751: An unresolved peak at 4.617 was assigned to the methine proton in the .CHI. group. The mass spectral results, so far obtained for (LXXXIV) are ambiguous in that the salt decomposed rapidly in the spectrometer and the mass peaks observed were those of the olefin(LXXXV).

CH .(CH2)8.CH:CH.CH3

Lluv This is not diagnostic of the dimeric nature of LXXXIV.

-77-

0) the reduction of the quaternary salt (LXXXIV) to the piperidine derivative (LXXXVI) was accomplished, and

H 2 /Pt0 2 LXXXIV >

CN.C-H.(CH-2,)tiCHI the molecular weight of this, determined by Rast's method, indicated the dimeric structure. This confirmed that the pyridinium salt (LXXXIV) was, in fact, the dieter. The n.m.r. spectrum of the piperidine showed a broad, unresolved band between 6.5t and 7.51 which could be assigned to the protons on carbons adjacent to the nitrogen. The only other absorption was the characteristic methylene absorption at 8.73`C and a distorted doublet at 9.l T. The latter absorption was assigned to the methyl protons on the Me.CH(NR2).CH2 group. The chloride (XI; x = Cl) was prepared by the following series of reactions:-

CH2:Ca(CH2)8CH2C1 C1CH2(CH2) . H2Br

II, LXXXVII

Nte, ( CH Cl 1)NaNH2/11q.NH3 .2 2). 2) LXXXVII 6N (231; x = Cl) -78-

The chloride (XI; x = 01) proved remarkably stable and was recovered unchanged after short path distillation. When the chloride (XI; x = C1) was heated in a large volume of high boiling solvent no quaternisation occurred. In a very high boiling solvent, such as 2-methylnaphthalene, thermal decomposition of the chloride occurred and no identifiable products were isolated. The bromide (XI; x = Br) was finally prepared by the following method:

CH (CH Me 2,___2)10.CH20Et 1)NaNH2/liq.NH3 25BrCH2,(CH2).901120Bt LXXXVIII

HBr

CH2OH2)10.CH2-Br

-,N,N,i(f±:x=Br) The primary nature of the bromide was confirmed by the n.m.r. spectrum, which showed a sharp triplet at 6.621: from the .CH2Br group. As with the secondary bromide (LXXXIII) no cyclic monomeric quaternary salt could be obtained. A salt, isolated by the procedure which had given the dimeric quaternary salt (LXXXIV) from the secondary bromide (LXXXIII), was assigned a similar -79- structure (LXXXIX). •cl-k 0 74 LXXXVIII Z

CI0A ) • c ‘11, 4 1 10 LXXXIX Each of the quaternary salts (LXXXIX and LXXXIV) showed in its n.m.r. spectrum absorption above 8.751: This absorption occurred in the form of three peaks at 8.95-r , 9.051E and 9.13T. The absorption in each spectrum corresponded to two protons for each pyridinium ring present in each molecule. It seems most unlikely that the two quaternary salts are, in fact, cyclic monomers. The evidence of the molecular weights does not suggest this and the n.m.r. spectra of the salts bear little resemblance to those of [10]- and [12]-paracyclophane (25). The simplest explanation is that the quaternary salts (LXXXIV and LXIXIX) are dimeric and that the spatial arrange- ment of the methylene chain relative to the pyridinium ring allows a part of this chain to spend some time in the diamagnetic shielding region around the aromatic ring. The absorption above 8.75T would then be the time- averaged effect of this spatial arrangement. It does not seem possible that the central methylene group, in the -80 - quaternary salts (LXXXIV and LXXXIX) are held rigidly above the aromatic ring as are those in a paracyclophane. Addition of Hydrogen Bromide to Olefins. In the synthetic work in this thesis, frequent use has been made of the addition of hydrogen bromide to olefins. Table 5 shows the various compounds which have been treated with hydrogen bromide and the nature of the bromide so obtained. All the olefins were terminal and the conditions of the reactions were identical. In each case a little dibenzoyl peroxide (ca. 1 mole %) was added to the reaction mixture. In experiments 4-10 (see over) the product is the primary bromide as predicted by Mayo and Welling (78). The mechanism of the reaction is predominantly free radical and the addition of peroxide to the reaction mixture is probably unnecessary. In experiments 1 and 2, the same type of internal radical inhibition, which was observed by Ashton and Smith (79) in the addition of hydrogen bromide to undec-10-en-l-ol, is probably operating. Undec-10- en-1-ol, even in the presence of peroxide, does not give the primary bromide. This correlates with the known mild radical inhibiting effect of primary alcohols. Amines, notably diphenylemine, exhibit a much stronger radical inhibiting effect. Thus, the tertiary amine, -81 -

TABLE 5

Olefin Bromide T00emp. Yield Solvent

(CH ) .CH:CH 1 2 Secondary 4 63% H2O and 01101 1,1 CN 3 2 H 0:011(0110).0H0 Secondary 4 - benzene 2 ' 8 ' li./',o 3 0113.00.0H(CH2)80H:OH2 mixture 4 68% it (prim.+sec,) 4 0H2:011.(CH2)80H201 primary 4 73% ft

5 CH2:011.(CH2)100H:C112 primary 4 70% tr Eta(CH2)ri0H:OH2

6 n = 9 primary 4 92% ft 7 n = 5 primary 4 64% ft 8 n = 3 primary 4 82% sc 9 0112:CH(CH2)8.0001 primary 4 82% If

10 OH2:011.(CH2)8.002H primary 4 62% 11

pyridine, may well exercise a similar effect and this correlates with the experimental results in Table 5. In the case of the ketone in experiment 3 the radical inhibiting effect of the carbonyl group must be so mild that the rate of the radical addition reaction is reduced - 82 - to about the same as that of the ionic addition reaction. Thus, a mixture of primary and secondary bromide is the product. - 83 -

CHAPTER 4 SPECULATIVE SYNTHETIC METHODS FOR BRIDGED PYRIDINES

Introduction. The ideas put forward in this chapter are unsubstan- tiated by any experimental work. They represent, in some cases, a completely different approach to the syn- thesis of bridged pyridines from those described in the previous chapter. In other cases they are adaptations of methods that have already been investigated. The most useful synthetic route to bridged pyridines will be the one which is applicable to the synthesis of other bridged aromatic systems. In particular, methylene bridged pyraz:ne, pyrimidine, and pyridazine systems should be capable of synthesis by this optimum route. These three heterocyclic aromatic rings are closely similar to pyridine in terms of molecular dimensions but the resonance energy of each one is apparently some 6-10 kcal./ mole. lower than that of pyridine. A comparison between the n.m.r. spectra of bridged pyridines, pyrimidines, pyridazines and pyrazines should clarify the relationship between ring currents and aromaticity. - 84 -

The Acyloin Condensation (32). a) The following route, involving a piperidine dicarboxylic ester, is suggested:

CH : C Hz c K Mel Olt (so) LXVCII

co cHl Br) (cHatIC011-1 0E1-0H -/ L-1 OH '7) ) HIIR*02. CO,i-1-(911,4N N/CH 4 Ph c‘S\c' b-`) "\)

This synthesis can be adapted to give a methylene bridge of a variety of lengths. The olefin (LXVIII) happened to be one whose synthesis has been described in this thesis. b) Muscopyridine is a 2,6-methylene bridged pyridine. Part of the synthesis of this natural product involves a ring expansion variation of the Schmidt reaction (82). -85-

pa jp_cvy„ne.

c_14,) ( /10

The pyridine (X0) was one of three products isolated. It was converted into muscopyridine by a further series of reactions. For the synthesis of a 2,5-bridged pyridine a bicyclic olefin (XCI) must be prepared

xdt The position of the double bond is unimportant as long as one and of the double bond lies on the bridge head carbon. Bredt's rule does not apply to a molecule such as XCI. The synthesis of the olefin (XCI) could be carried out in the following manner:-

- 86 -

/ CH ; \//C.H CH Nick.CH(co,O, e 0 ?) \-Iy6-oly is 3) —col_

C 0, (C.1-Cz) Hz 014-141.1 /1-2. C

6 H, CH2., CH, OTs 1) L. A. (00 , 0‘< Br 1-)Ts CH2- c1;1,

cH . CH4. C 1-‘2 43r

us —H2O Co

- 87 -

Friedel-Crafts! Reaction.

The intramolecular adaptation of the Friedel-Crafts, reaction may be used to prepare bridged pyridines in the following synthesis:-

cH: Kt\mi/tivlo-1,

Lxvt (5 ),cH,cL ( 1 )

. CH :-C1-12. 9 KMn0 4. (C SOC12.

• COCL 4- @Hi Ni H CL

Ro y\et N; H -2) 1,01- Kskner CO -88-

Diacetylene Method. The work on the synthesis of diacetylenes may not be entirely wasted. Bridged pyridines may be prepared from acetylenes of the type (LXX) by the following method (50):

C:CH N N1421Iii.. (C H2}„ B r

The reaction would be run in liquid ammonia at high dilution. -89 -

CHAPTER 5

EXPERIMENTAL

Infra-red spectra were recorded on a Unlearn S.P.200 spectrometer. N.m.r. spectra were recorded on a Varian A.60 spectro- meter. The tau (t) scale, where tetramethylsilane is assigned a value of 10, is used throughout. The chemical shifts are quoted in the text as T, chemical shift (multiplicity, assignment). Intensities are unspecified, being consistently correct for the assignments made. All the n.m.r. spectra were recorded for solutions in carbon tetrachloride except those marked 1:3. where deutero- chloroform was the solvent. The abbreviations used are: S = singlet, D = doublet, T = triplet, Q = quartet, M = multiplet, U = unresolved, (0112)n = chain of methylene groups containing all such groups not otherwise assigned, Ar. = aromatic ring, usually pyridine. All melting points were determined on a Kofler black and are corrected. Thionyl thioride was always redistilled from quinoline before use. The term "dry ether" refers to sodium-dried ether. Hydrogen bromide was prepared by treating dry • -90- tetralin with bromine. The gas, thus obtained, was passed through two wash-bottles containing dry tetralin before use. The term ttoetroleum ether" refers to the petroleum fraction boiling between 600 and 80°. Thin layer chromatograms were eluted with chloroform. The absorbent was silica gel H.P.254. These chromato- grams were particularly useful in the analysis of products from the attempted coupling reactions. 91 -

The 2,5 Bridged Pyridine

Experimental Section -92-

6-Methylpyridine-3-carbinol (XIX). Ethyl 6-methylnicotinate (6 g.), dissolved in dry ether (39 mis.), was added to a stirred slurry of lithium aluminium hydride (1.8 g.) in dry ether (78 mis.) over 40 minutes. The ether refluxed gently. The addition completed, stirring was continued for 30 minutes and then the excess of hydride was decomposed by ethyl acetate. Water (5 mis.) was added and the mixture filtered. The filter-cake was suspended in methanol (50 mis.) and carbon dioxide was bubbled through. The suspension was boiled briefly and filtered hot. The solid was treated likewise a second time. The combined methanolic and ethereal filtrates were evaporated and the residue dissolved in ether (50 mis.). The ethereal solution was dried (K2003), filtered and evaporated. The residue was distilled under high vacuum. Yield: 2.6 g. (59%), b.p. 80.06 mm. (liquid quiescent). 3 no 1.5570. Calc. for 0 H N0: (Found: 0, 68.7; H, 7.4; N, 11.3. 7 9 0, 68.3; H, 7.3; H, 11.0.) 17, 7.62 (S,Me), 5.54 (S, -0112.0H), 4.70 (S1-011), 3.04 (D, 2.36 (Q, Y-H), 1.89 (D,m-H). -93-

6-Methy1-3-pyridylmethyltoluene-2-sulphonate (XIII) .

Several attempts were made to prepare this compound from 6-methylpyridine-3-carbinol. Generally, the instructions of Tipson (55) were followed. 6-Methylpyri- dine-3-carbinol (2 g.) was added to a solution of R7toluene- galphonyl chloride (3.4 g.) in dry pyridine (13.5 mis.) held at -5°. The mixture was then allowed to stand at 0° for 2 hours. The work-up procedure was varied con- siderably. An aqueous work-up followed by extraction with chloroform was used without success. Evaporation of the pyridine under reduced pressure followed by recrystallisation of the residue was equally unsuccessful. Addition of water (2 mis.) to the mixture followed by sodium bicarbonate until effervescence ceaseda-ld evaporation of the solvent yielded no tosylate. 6-Methylpyridine-3-carbinol was recovered unchanged. It could be isolated as a picrate, m.p. (from ethanol)

182°. (Found: C, 44.9; H, 3.1. 013H11N408 requires C, 44.5; H, 3.1%.) 2-Meth 1-5- 1-oxododec-11-en 1 • ridine (XXI) (57). 6-Methylnicotinonitrile (29.5 g.), in dry ether (700 mls.),was added during 1 hour to an ice-cooled solution of undec-10-enylmagnesium chloride, prepared from -94-

1-chloro-undec-10-ene (110 g.) and magnesium (11.4 g.), in dry ether (200 mis.). The addition completed, the mixture was stirred at 0° for 30 minutes, then at 20° for 2 hours and finally under gentle reflux for 2 hours. The mixture was cooled to 0° and a saturated aqueous solution of ammonium chloride (225 mis.) stirred in. This was followed by concentrated hydrochloric acid (70 mis.). The acidic mixture was stirred at 20° for 30 minutes and then treated with 4N sodium hydroxide solution (250 mis.). The ethereal layer was removed and the aqueous layer extracted with ether (4 x 250 mis.). The combined ethereal

solutions were dried (Na2SO4), filtered and evaporated. The residue was distilled under high vacuum. Yield: 37 g. (56%). Short path distillation at 120°/5 x 10-6 mm.

(Found: 0, 79.5; H, 10.3; N, 5.0(5). C18H27NO requires 0, 79.1; H, 10.0; N, 5.1%.)

max 1687 (0=0), 1640 0=0), 1598, 1565, 1495 (Ar.)cm.-1 17, 8.70 (S,(CH2)n), 7.45 (S,MeAr.), 7.12 (T,CH2.00), ca.5.l (M,CH2=), ca. 4.4 (M, CH=.), 2.85 (D,B-H), 2.01 (Q, Y-H), 1.13 (D,m-H). By the above method the following pyridines were prepared. 2-Methy1-5-(1-oxo-8-ethoxyoctyl)p2yridine CXXX . This was prepared from 6-methylnicotinonitrile (17 g.), -95-

7-bromo-l-ethoxyheptane (57.3 g.) and magnesium (4.9 g.). Yield: 28 g. (73%). b.p. 140-145°/.07 mm. 1680 (C=0), 1600, 1561, 1495 (Ar.), 1110 (ether) CM. -1 The infrared spectrum showed traces of nitrile (2210 m N-) cm.-1) ir , 8.88 (T,C43.0112.0), 8.62 (S,(CH2)n), 7.44 (S, MeAr.) 7.11 (T,CH2.00), 6.68 (T,CH2.CH2.0), 6.62 (Q,CH3OH2.0) 2.85 CD, P-H), 2.01 (Q, Y-H), 1.12 (D, a-H). Oxime: m.p. (from petroleum ether ) 90°. (Found: Cy 69.6; H, 9.4; N, 10.5. 16H26N202 re',uires CI 69.0; H, 9.4; N, 10.1%.)

This was prepared from 6-methylnicotinonitrile (20 g.), 4-bromo-l-ethoxybutane (55 g.) and magnesium (7.3 g.). Yield: 18 g. (48%). b.p. 150°/.9 mm. 1680 (C=0), 1600, 1562, 1495 (Ar.), 1110 (ether) cm.-1 . 8.88 (T, CH3.CH2.0), ca. 8.3 (M,(0112)n), 7.44 (SyMe.Ar.), 7.11 (TyCH2.00), 6.64 (T, CH2.0112.0.), 6.63 (Q, CH3.CH2.Ar.), 2.85 (D, 3-H), 2.01 (Q, Y-H), 1.13 (D, a-H). Picrate: m.p. (from ethanol) 110.8°.

(Pound: Oy 50.6; H, 5.2; N, 12.6. 0191122N409 requires 0, 50.6; H, 4.9; N, 12.4%). -96-

2-Methy1-5-(1-oxo-he7-5-enyl)pyridine (LXV). This was prepared from 6-methylnicotinonitrile (31.6 g.), 1-chloropent-4-ene (56 g.) and magnesium (13.7 g.). Yield: 23 g. (46%). b.p. 106°/.5 mm. max. 1685 (C=0)1 1638 (C.C), 1596, 1563, 1495 (Ar.) am.-1. 5 ca. 8.0 (Mt (0112)n), 7.46 (St Me.Ar.), 7.12 (TICH2.00), ca. 5.1 (Mt CH2=), ca. 4.5 (Mt CH=), 2.86 (S, 0-H), 2.01 (Qt Y-H), 1.13 (D, a-H). picrate: m.p. (from ethanol) 104-106°. wound: N, 13.4. 018H18N308 requires N, 13.4%). 5-Dodec-11'-eny1-2-methylpxridine (XXII) (57). 2-Methyl-5-(l-oxododec-11-enyl)pyridine (10 g.), hydrazine hydrate (5.7 g.) and potassium hydroxide (4.6 g.) were dissolved in triethylene glycol (60 mis.). The mixture was stirred at 120° for 1 hour, then at 160-180° for 2 hours and finally at 205° for 15 minutes. The cooled mixture was diluted with water (100 mis.) and ex- tracted with benzene (3 x 100 mis.). The combined benzene solutions were dried (112003), filtered and evaporated. The residue was distilled under vacuum. Yield: 7.0 g. (74%). b.p. 107-112°/.001 mm. n145-6 1.4894. (Pound: 0, 83.4; H, 11.4; N, 5.1. C18H29N requires CI 83.3; H, 11.3; N, 5.4%.) :\1u max. 1638 (C=0), 1601, 1565, 1495 (Ar.) cm.-1. -97-

8.73 (Sp(CH2)n), 7.53 (St Me.Ar.), 7.48 (T, CH2.Ar.), ca. 5.1 (M, CH2.)p ca. 4.4 (M, 01-1=), 3.11 (D, p-10, 2.79 (Q, Y-11), 1.75 (D, m-H). By the above method the following pyridines were prepared. 5-8'-Ethoxyocty1-2-methylpyridine (XXXI). This was prepared from the ketone (25 g.), hydrazine hydrate (14.7 g.) and potassium hydroxide (11.8 g.) in triethylene glycol (154 mls.). Yield: 17.6 g. (75%). b.p. 100-110°/.2 mm. n21 1.4724. (Found: N, 5.1. C16H27N0 requires N, 5.6%.) max. 1600, 1565, 1495 (Ar.), 1110 (ether) cm.-1. 1[1 8.88 (T, 0H3.0112.0), 8.72 (Sp (0H2)n), 7.58 (S, Me.Ar.), 7.50 (T, CH2.Ar.)p 6.70 (Tp 0112.0112.0), 6.64 (Q, CH3.CH2.0), 3.20 (D, p-1), 3.85 (Q, Y-H), 1.80 (D, m-H). 5-5,-Ethoxypenty1-2-methylvridine (XXXVIII). This was prepared from the ketone (15 g.), hydrazine hydrate (10.7 g.) and potassium hydroxide (8.5 g.) in triethylene glycol (150 mis.). Yield: 10 g. (71%). b.p. 80-82°/0.1 mm. n23 1.4859. (Found: 0, 75.0; H, 10.2(5). C13H21N0 requires Cp 75.3; H, 10.3%4) -98-

L3 max. 1605, 1565, 1495 (Ar.), 1110 (ether) cm.-1.

8.88 (T, CH3.CH2.0), 8.55 (S, (CH2)21), 7.56 (S, Me.Ar.), 7.50 (T, CH2.Ar.), 6.70 (T, 0112.0H2.0), 6.64 (Q, CH3.0112.0), 3.11 (D, (3-H), 2.80 (Q, Y-H), 1.75 (D, m-H). 5-Hex-5'-eny1-2-methylpyridine (LXVI). This was prepared from the ketone (21.5 g.), hydra- zine hydrate (18 g.) and potassium hydroxide (14.4 g.) in triethylene glycol (188 mis.). Yield: 14 g. (71%). b.p. 118-123°/16 mm. 1 n0 1.5022. (Found: 0, 81.7; H, 9.8; N, 7.8. 012 1117N requires 0, 82.2; H, 9.8; N, 8.0%.) "max. 1638 (0=0), 1600, 1565, 1495 (Ar.) cm.-1. 7.8-8.6 (M, (0H2 )n, 7.55 (8, Me.Ar.), 7.50 (T, CH2.Ar.), ca. 5.1 (M, 0H2=), ea. 4.4 (11, CH=), 3.10 (D, i3-H), 2.79 (Q, Y-H), 1.75 (D, m-H). -99-

24.11,12-Dihydroxydodecy1)-2-methylpyridine (Mal (80,86). This was prepared by treating 5-dodec-11,-eny1-2- methylpyridine (3.9 g.) with silver benzoate (6.9 g.) and iodine (3.8 g.) in dry benzene (75 mis.) according to the method of Brody and Bogert. Yield: 1 g. (230). m.p. (from petroleum ether) 57-550. (Found: C, 74.0; H, 10.8; N, 4.4. 01031NO2 requires 0, 73.7; H, 10.7; N, 4.8%. The yield was not increased by using freshly prepared silver benzoate and running the reaction in the dark. 17, 8.68 (s, (CH2)n), 7.48 (S, Me.Ar.), 7.44 (T, CH2.Ar.), ca. 6.3 (M, CHOH.CH2OH), 2.90 (D, p-H), 2.57 (Q, Y-H), 1.66 (D, m-H). 5-101 -Formyldecy1-2-methylpyridine (XII). 5-(11,12-Dihydroxydodecy1)-2-methylpyridine (0.1 g.) was shaken with water (1.2 mis.) and the pH of the mixture adjusted to 4 with dilute sulphuric acid. Periodic acid (0.1 g.) in water (1.2 mis.) was added and the mixture shaken for 20 hours. The mixture was made alkaline with -100-

barium hydroxide solution and filtered. The filter cake was washed with chloroform (20 mis.) and the filtrate extracted with chloroform (2 x 20 mis.). The combined chloroform solutions were dried (Na2SO4)' filtered and evaporated. The residue (ca. 10 mg.) was examined spectroscopically. )) max. 1708 (0=0), 1603, 1570, 1495 (Ar.), 865 (C-H) cm.-1. 2-3f-EthoxYpropy1-5-etivlpyridine (XXXII) (59). A solution of phenyl-lithium in dry ether (250 mis.) was prepared from lithium (5 g.) and bromobenzene. 5-Ethyl-2-methylpyridine (41.6 g.) in dry ether (50 mis.) was added to this stirred solution during 30 minutes at 200. No cooling was necessary. Then, with vigorous stirring, 2-chloroethyl ethyl ether (19.5 g.) in dry ether (50 mis.) was added rapidly. The addition produced gentle refluxing and this was maintained for 30 minutes. The mixture was cooled and poured onto ice (500 g.). The ethereal layer was removed and the aqueous layer extracted with ether (2 x 100 mis.). The combined ethereal solutions were dried (K2CO3), filtered and evaporated. The residue was distilled under vacuum. 25 Yield: 2.5 g. (73%). b.p. 107-110/.6 mm. n 1.4979.

(Found: N, C, 74.5; H, 9.7; 7.2. C12H19N0 requires

0, 74.6; H, 9.9; N, 7.2%a) -101 -

%max. 1600, 1565, 1494 (Ar.), 1110 (ether) cm.-1. Ty 8.85 (T, 0H3.0H2.0.), 8.76 (T, Ar.0H2.0H3), ca. 8.1 (My 0.0H20H2.0112.Ar.), ca. 7.3 (My Ar.0H2), 6.62 (T, CH2.0112.0.), 6.62 (Q, cli3.q2.0), 3.04 (D, (3-H), 2.69 (Q, y-H), 1.72 (D, ac-H) . 2..2-t-Ethw:rrol-5-8t-ezg coctlridinenDEVI (59). This was prepared from 2-methy1,5-81 -ethoxyoctyl- pyridine (16 g.) and 2-Chloroethyl ethyl ether (6.9 g.). The ratio PhLi:base:halide was 1:1:1 instead of 2:2:1 as above. Yield: 6 g. (30%). b.p. 150°/.2 mm. n18 1.4831. (Pound: 0, 74.4; H, 11.0; N, 4.5. C20H35 02 requires 0, 74.7; H, 11.0; N, 4.4%.) k)max. 1600, 1565, 1495 (Ar.), 1110 (ether) cm. -1. T 8.86 (T, CH3.0H2.0), 8.67 (Sy(CH2)n)y 7.37(TO-CH2.Ar.) 7.23 (T, m-CH2.Ar.), ca. 6.6 (MI CH2.0.CH2), 3.02 (D, P-H), 2.66 (Q, Y-H), 1.72 (D, a-H). 2-61Ethoxyhexy1-5-51 -ethoxypentylpyridine (XXXIX) (59). This was prepared from 2-methyl 5-5g-ethoxypentyl- pyridine (7.5 g.) and 1-bromo-5-ethoxypentane (4.9 g.) by the above method. Yield: 5 g. (43%). b.p. 145°/.15 mm. (Pound: Cy 74.6; H, 10.8; N, 4.7. 020H35NO2 requires Cy 74.7; H, 11.0; N, 4.40.). 102 -

S liaax. 1600, 1566, 1492 (Ar.), 1110 (ether) cm.-1. 8.87 (T, CH3CH2.0), 8.54 (s, (CH2)n), 7.43 (T, P-CH2.Ar.) 7.31 (T, a-CH2.Ar.), 6.68 (T, CH2.0112.0), 6.65 (Q, CH3 .CH 0)1 3.05 CD, (3.-H), 2.79 (Q, T•41), 1.75 (D, a-H). 2-61 -Ethoxyhexy1-5-ethyltEridine (XL). (59). This was prepared as above from lithium (2.2 g.), bromobenzene (24.3 g.), 5-ethyl-2-methylpyridine (18.7 g.) and 5-bromo-l-ethoxypentane (15 g.). Yield: 13 g. (65%), b.p. 84°/.05 mm. 1.4915. (Pound: N, 5.8(5). C15H25N0 requires N, 5.9(5)%.). %302,7.. 1601, 1564, 1494 (Ar.), 1110 (ether) cm.-1. 7E, 8.85 (T, CH3.CH2.0), 8.76 (T, CH3,0112.Ar.), Ca. 7.4 (M, Ar.CH2), 6.68 (T, CH2.CH2.0.), 6.65 (Q, 01-13.U2.0.) 3.04 (D, 0-11), 2.69 (Q, y-11), 1.72 (D, a-H). 2-31 -BromoyroRy1757ethdy2...33 0EVM (87) . 2-3,-Ethoxy-propy1-5-ethylpyriline (14 g.) was dissolve in 60% aqueous hydrobromic acid (30 mis.) and the solution was cooled to 0° in a Carius tube. The solution was then saturated at this temperature with hydrogen bromide and the Carius tube sealed. The tube was heated at 120° for 2 hours. The tube's contents were evaporated under reduced pressure and the residue was treated with charcoal in acetone. The filtered acetone solution was evaporated to 70 mis., cooled and filtered. The crystals were washed with ether. 4:103 -

Yield: 12 g. (52%). m.p. (from NV/acetone) 1300. (Found: C, 39.0; H, 4.4(5); Br 51.8. C10H5Br2N requires 0, 39.2; H, 4.8(5); Br, 51.7%.) 17, 8.64 (T, CH3.CH2), 7.45 (Mt Br.CH2.CH2.CH2.Ar.), 7.02 (Q, 0H3.CH2 ), 6.55 (T, CH2 .CH2.Ar.), 6.48 (T, CH2.33r), 2.09 (D, 3-H), 1.55 (Q, Y-H), 1.05 (D, a-H). Attempted preparation of 2-31-cyanopropy1-5-ethylpyridine (XXXIV) (61). (1)A solution of sodium cyanide (2.65 g.) in water (4 mis.) was mixed with a solution of 2-31-bromopropy1-5-ethylpyridine (8 g.) in ethanol (16 mis.). The mixturecwas refluxed for 8 hours, allowed to cool and then diluted with water (100 mls.). The aqueous solution made alkaline with 4N sodium hydroxide solution and extracted with ether (4 x 100 mis.). The combined ethereal solutions were dried (X2003), filtered and evaporated. They contained no organic product. (2)The hydrobromide (4 g.) was added to a refluxing mixture of auprous cyanide (1.5 g.) and phosphorous pentoxide (0.2 g.) in acetonitrile (20 mis.). The mixture was refluxed for 8 hours. The work up and result were similar to (1) above. (3)The hydrobromide (5 g.) and sodium cyanide (1.33 g.) were added to a solution of hydrogen cyanide (2 mis.) in - 104 -

80% aqueous ethanol (50 mls.). The solution was refluxed for seven haurs. The work up and result were similar to (1) above. In each case the organic material appears to have been retained in the alkaline aqueous phase. Ethyl undec -10 -enoate (XLVII) (76). This was prepared by the esterification of undec -10 - enoic acid (405 g.). The method was that of Barkovsky. Yield: 434 g. (93%). b.p. 1310/16 mm. n23 1.4378. Lit. b.p. 1310/14 mm. max. 1715 (0=0), 1638 (0=0) cm.-1 . Undec -10 -en -1 -01 (XLVIII) (76). a) A slurry of lithium aluminium hydride (6.8 g.) in dry ether (200 mls.) was prepared. The slurry was stirred and ethyl undec -10 -enoate (51 g.) in dry ether (200 mis.) was added at a rate such that the ether refluxed gently. The addition completed, the excess of hydride was decomposed with methyl acetate. The resulting, very viscous mixture was poured onto ice (200 g.) and acidified with 4N sulphuric acid. The ethereal layer was removed and the aqueous layer extracted with ether (2 x 100 mls.). The combined ethereal solutions were washed with saturated sodium carbonate solution (100 mis.) and water (100 mis.). The dried (MgSO4), filtered, ethereal solution was evapor- ated and the residue distilled under vacuum. -105-

Yield: 36.8 g. (90%). b.p. 95°/.l mm. nil 1.4492. Lit. b.p. 122°/3 mm. n19 1.4502. b) The reduction was also performed according to the method of gran and Wirth using sodium and anhydrous ethanol. The yields, on two separate runs, were 65% and 70%. The physical constants of the alcohol were as above. 1-Ghloroundec-10-e2214 Lal (76). This was prepared by treating undec-10-en-1-ol (277 g. with thionyl chloride (243 g.). p. 750/0.5 mm. Yield: 236 g. (93%). b. 49.5 1.4502. Lit. b.p. 113--115°/l2 mm. nn4 = 1.4926. max. 1638 (0=0) an.-1. 8.68 (S,(CH2)n), 6.55 (T, 0H201), ca. 5.1 (Mt =0H2), ca. 4.4 (M, -01.1=). Undec-ID-eTayl chloride (q1 (88). This was prepared from undec-10-enoic acid (150 g.) by treatment with thionyl chloride (205 g.). Yield: 144 g. (87%). b.p. 114-115/5 mm. Lit. 128°/13 mm. max. 1790 (C=0), 1639 (C=0) cm.-1. 11-Bromoundecanoic acid (79). This was prepared from undec-10-enoic acid (42.0 g.), by treatment in benzene solution with hydrogen bromide 106 -

according to the method of Ashton and Smith. Yield: 40 g. 62%. m.p. (from petroleum ether) 50°. Lit. m.p. 49°. 1[329 8.68 (S (CH ) )9 7.68 (T, H .0O2H), 6.64 (T, CH2Br), -2.38 (8,-002M). 11-Bromoundecanoyl chloride (LII). Undec-10-enoyl chloride (144 g.), dissolved in benzene (I1.) was treated with hydrogen bromide as in the prepara- tion of 11-bromoundecanoic acid. Yield: 200 g. (98%). This acid chloride could not be distilled without decomposition and was a liquid above 2°. However, if a little of it was warmed with water (10 mis.), an acid was obtained. This acid was identical (mixed melting point) with the 11-bromoundecanoic acid obtained above.

(For acid: Found: 0, 49.9; H, 7.9. Oalc.for C11H21Br02: C, 49.8; H, 8.0%.) The acid chloride was also made by treating 11-bromo- undecanoic acid (38 g.) with thionyl chloride (32.7 g.) as for undec-10-enoyl chloride. The bromo acid chloride, prepared by this method, was considerably less pure than that prepared as above. 11-Bromoundecan-l-ol. LSIT (89). Crude 11-bromoundecanoyl chloride (31.6 g.) was dissolved in dry ether (350 mis.). To this stirred solution, kept at 001 a solution of lithium aluminium - 107 - hydride (165 mis. containing .37 moles L.A.H. per litre) was added at a rate such that the tcmperature of the solution was kept below 5°. When the addition was complete the excess of hydride was decomposed with ethyl acetate (5 mis.). Water (100 mis.) was addedp follawed by 2N sulphuric acid (75 mis.). The ethereal layer was removed and the aqueous layer extracted with ether (2 x 50 mis.). The combined ethereal solutions were washed with 10% sodium bicarbonate solution and water. The ethereal solutions was dried (MgSO4), filtered and evaporated. The residue crystallised when cooled. Yield: 20.5 g. (73%). m.p. (from petroleum ether) 45°. Lit. m.p. 44. (Found: C, 52.4; 8, 9.0. Cale. for C11H23Br0:0, 52.6; H, 9.3%.) TR, 8.70 (S,(0112)n), 7.86 (S,OH), 6.67 (T, CH2 .Br), 6.48 (T,CH2.0H). 12-Hydroxydodecanonitrile (LIV (90). This was prepared from 11-bromoundecan-l-ol (120 g.) by treatment with potassium cyanide (34 g.) in 87% aqueous ethanol. Yield: 75 g. (C0%). b.p. 158-162°/.5 mm. m.p. (from CC14/petrol'.um ether) 35.50. (Pound: C, 73.0; H, 11.8; N,7.2. C12112 3NO requires C, 73.0(5); H, 11.7(5); N, 7.1%.) -103-

max. 2220 (C=T) cm.-l. 2, 8.70 (S,-(CH2)n)1 7.80 (8,-OH), 7.66 (T,-CH2CN), 6.40 (T, CH2OH). Dodec-11-enonitrile (L) (91). 1-Chloroundec-10-ene (37 g.) was treated with potassium cyanide (35 g.) in 86% aqueous ethanol (518 mis.) to yield the nitrile (L). Yield: 21 g. (60%). b.p. 86-92°/.9 mm. ne 1.4465. Lit. b.p. 145°/15 win. t) max. 2230 (0.4."'1 2 1640 (C=C) am.-1. Tridec-12-en-2-one (XLIV; n=10 (91). This was prepared by treating dodec-11-enonitrile (20 g.) with methyl magnesium iodide, prepared from magnesium (11.5 g.) and methyl iodide (62 g.).

Yield: 7.5 g. (34%). b.p. 76°/.5 mm. n018 1.4466. Lit. b.p. 133713 mm. 0 max. 1713 (C=0), 1639 (0=0) cm.-1. 2:4 D.N.P. m.p. (from ethanol) 63-64°. (Found: 0, 60.8; H, 7.6. C19H28N404 requires C, 60.6; H, 7.5%.) Dodec-ll-en-2-one (XLIV; 11E21 (91,92). This was rrepared by treating dimethyl-cadmium, prepared from magnesium (19.1 g.), methyl bromide (71 g.) and cadmium chloride (72.3 g.), with undec-10-enoyl chloride

-109 -

(169 g.). Yield: 57 g. (40%). b.p. 108-120°/8 mm. Lit. b.p. 114°/9 mm. \) max. 1713 (0=0), 1640 (C=0) cm.-1. 13-Hydroxytridecan-2-one. (XLV) (93). This was prepared by treating 12-hydroxydodecano- nitrile (15.7 g.) with methyl magnesium bromide, prepared from methyl bromide (48.5 g.) and magnesium (11.5 g.). Yield: 13 g. (76%). m.p. (from petroleum ether) 54-56°. (Pound: 0, 73.4; H, 12.5(5). 013H2602 requires C, 72.9; H, 12.2%.) -1 max. 1708 (C=0) an. . , 8.73 (S,-(CH2)n-)9 7.86 (S,CH3.00), ca. 7.58 (U, -OH —cE2ocio), 6.38 (T, -042.0H). 2:4 D.N.P. m.p. (from methanol) 75-76.5°. (Found: N, 14.5. 019H31N403 requires N, 14.2%.) 13-Triphenylmethoxytridecan-2-one (LVIII; n=10) (64). 13-Hydroxytridecan-2-one (20 g.) and triphenylmethanol (24.4 g.) were dissolved in dry benzene (370 mis.). p-Toluene sulphonic acid (0.5 g.) was added to the solution and the mixture was refluxed. The water, which was produced in the reaction, was collected in a Dean and Stark apparatus. When the evolution of water ceased, the cooled solution was washed with 5% sodium bicarbonate -110 -

solution (100 mis.) and then with water (100 mis.). The benzene solution was evaporated under reduced pressure. The residue was a dark,red oil. This oil was examined spectroscopically. 1692 (C=0), 1580, 1480 (Ar.), 1200 (OAr.), 1140 (OAlk.) an.-1. T, 8.75 (S,(CH2)n), 8.07 (s, CH3.00), 7.76 (T, CH2.00.),

6.98 (T, CH2.0.), ca. 2.7 (N, Ar.-H). The n.m.r. spectrum of a deuterated sample was identical to that given above. 3-Cyano-6-undec-109 -enyl-2-pyridone (LVI) (62). Sodium sand (0.8 g.) was covered with dry ether (50 mis.) and the mixture cooled to 00. A mixture of ethyl formate (2.6 g.) and tridec-12-en-2-one (7 g.) was added over 1 hour. The mixture had to be warmed at first to start the reaction. The addition completed, stirring was continued for 2 hours at 200. The solvent was then evaporated under reduced pressure and the residue dissolved in water (15 mis.). Cyanoacetamide (2.1 g.) was added followed by a mixture of glacial acetic acid (.34 mis.) in water (.85 mis.) rendered alkaline with piperidine. The mixture was refluxed for 2.5 hours, cooled and acidified with glacial acetic acid. A red oil separated and this rapidly crystallised. Yield: 1.2 g. (13%). m.p. (from petroleum ether) 94-96°. (Found: C, 75.2; H, 9.1; N, 10.3. C17H24N20 requires Cy 74.9; H, 8.9; N, 10.3%.) max. 2210 (CmN), 1660 (C = 0, 1640 (C=0), 1605, 1575, 1495 (Ar.) cm.-1 , 8.67 (8, (0H2)n), 7.56 (8, Me-Ar.), 7.28 (T, CH2.Ar.), ca. 5.06 (M, =CH2), ca. 4.3 (11,-0H=), 3.84 (D, P-H), 2.36 (8, Y-H), 2.24 (D, Y-H). For a discussion of a similar spectrum see Chapter 3. 3-0vano-6-dec-9t-eny1-2-pyridone (LV) (62). This was prepared by the above method from sodium (1.3 g.), ethyl formate (4.2 g.) and dodec-11-en-2-one (10.3 g.). Yield: 2 g. (14%). m.p.(from petroleum ether) 86-88°. (Found: C, 74.0; H, 8.4; N, 10.5. C16H22N20 requires C, 74.3; H, 8.6; N, 10.8(5)%.) 1:1% 8.67 (S,(CH2)n), 7.56 (S,Me-Ar.), 7.28 (T, CH2-Ar.)'1 ca. 5.06 (M,=CH2), ca. 4.3 (DI, -0/1=), 3.84 (D, c3-H), 2.36 (8, y-H), 2.24 (D, Y-H). For a discussion of this spectrum see Chapter 3. Hydrobromjmation of 3-c ano-6-dec-9f-enY1-2-yirridone. 3-Cyano 6-dec-9t-eny1-2-pyridone (0.7 g.) was treated with hydrogen bromide in the presence of perL.cide as before. Evaporation of the solvent yielded a red oil -112 -

which did not crystallise. Its n.m.r. spectrum showed that a secondary bromide had been formed. , 8.68 (S,(CH2)n-), 8.32 (D, CH3.CHBr.), 7.62 (T, -CH2.Ar), 5.94 CH3.CHBr), 3.83 (D, 0-H), 2.22 (D, Y-H), The n.m.r. spectrum also contained small peaks at 7.561' (SI CH3-Ar) and 2.34-C (S, Y-H). These are from the isomeric pyridone, 3-cyano 5-methyl 6-non-8I-enyl 2-pyridone. See Chapter 3. Eydrobromination of dodec-11-en-2-cne. Dodec-11-en-2-one (16 g.), dissolved in benzene (300 mis.), was treated with hydrogen bromide in the presence of dibenzoyl peroxide. Yield: 15.5 g. (68%). b.p. 102-119°/0.2 mm. 22 n 1.4671 to 1.4685. max. 1708 (0=0) an.-1 The n.m.r. spectrum of this bromoketone showed that a mixture of primary and secondary bromides had been formed. They were not separable by distillation. I-, 8.70 (S,(CH2)n), 8.28 (D, CH3.CHBr), 7.94 (8, CH3C0), 7.64 (T, CH2C0), 6.64 (T, CH2Br), ca. 5.9 (M, CH3.CHBr). Tetradeca-1,13-diene (LIX) (94). Undec-10-enyl magnesium chloride was prepared from the chloride (175 g.) and magnesium (22.2 g.) in dry ether (500 mis.). The Grignard solution was cooled to -113 •

0° and treated at this temperature with a solution of allyl bromide (110 g.) in dry ether (400 mis.) with vigorous stirring. The addition completed, the mixture was stirred for 30 minutes without cooling and then refluxed for 10 hours. The cooled mixture was poured onto ice (500 g.) and made acid with dilute hydrochloric acid. The ethereal layer was removed and the aqueous layer extracted with ether (2 x 100 mis.). The combined ethereal layers were dried (0a012)1 filtered and evaporated. The residue was distilled under vacuum. n21.5 Yield: 146 g. (81%). b.p. 75-76°/.3 mm. 1.4439. Lit. b.p. 131°/17 mm. n25 1.4427. *N) max. 1638 (0=0) an.-1. 1,14-Dibromotetradecane (L11 (95). Tetradeca-1,13-diene (136 g.) was dissolved in dry benzene (1.2 1.), containing a trace of dibenzoyl peroxide. Into the solution, which was kept at 4°, a stream of bromine-free hydrogen bromide was passed. When hydrogen bromide was no longer absorbed, the passage of gas was discontinued. The benzene was evaporated under reduced pressure and the residue cooled. Yield: 174 g. (70%). m.p. (from ethanol) 47-49°. Lit. m.p. 50.4°. Hexadecano-1,16-dinitrile (LXI) (96,97). 1,14-Dibromotetradecane (170 g.) was dissolved in ethanol (480 mis.) and this solution mixed with a solution of potassium cyanide (68 g.) in water (68 mis.). The mixture was refluxed for 50 hours, cooled and diluted with water (500 mis.). The aqueous solution was extracted with ether (1 x 200, 2 x 100 mis.) and the combined ethereal solutions dried (Na2SO4)' filtered and evaporated. The residue was distilled under vacuum. Yield: 70 g. (65%). m.p. 47-49°. b.p. 160-170°/.3 mm. Lit. m.p. 48-49°. b.p. 216-2200/4 mm. max. 2220 (C11-11) cm.-1. Cyclopentadecanone-Ziegler Method (45). A mixture of dry benzene (500 mis.), sodium wire (26.6 g.) and ohlorobenzene (68 g.) was stirred at room temperature under dry, oxygen-free nitrogen. The for- mation of phenyl-sodium proceeded smoothly. Gentle cooling was required to keep the temperature below 40°. The reaction was complete after 12 hours stirring. To the solution of phenyl sodium, kept at 0°, a solution of monoethylaniline (70 mis.) in dry ether (200 mis.) was added with stirring over 30 minutes. Stirring was con- tinued for 4 hours at 20°. The mixture was then filtered into the reaction vessel and the filtrate diluted with dry ether (1.5 1.). -115 -

To this stirred, refluxing solution, hexadecano- 1,16-dinitrile (8.5 g.) in dry benzene (35 mis.) was added continuously over 3.5 days. The addition complete, the solution was diluted with water (700 mis.). The ethereal layer was removed and evaporated. The residue was treated with 70% sulphuric acid (214 mis.) at 130° for 5 hours. The cooled mixture was diluted with water (500 mis.) and steam distilled. The organic material in the steam distillate was extracted with ether (3 x 50 mis.). The combined ethereal solutions were dried (MgSO4), filtered and evaporated. The residue was a white crys- talline solid. max. 1705 (0=0), 1600, 1490 (Ar) cm.-1. The residue was chromatographed on alumina (Type H). The aromatic material was biphenyl. m.p. 69-72°. Lit. 70°. (Found: C, 93.4; H, 6.6. Calc. for C12H10 C, 93.5; H, 6.5.) )max . 1600, 1490 (Ar.) cm.-1 . The ketonic material was cyclopentadecanone m.p. 60-64°. Lit. 63°. max. 1705 (C=0) cm.-1.

Yield: 0.7 g, (9%). -116 -

In another run, the lithium salt of monoethylaniline was used. Here the yield of steam volatile material was largely biphenyl. The yield of cyclopentadecanone was not improved by using concentrated hydrochloric acid instead of 70% sulphuric acid in a run using the sodium salt of mono- ethylaniline. In an exploratory experiment using sodium hydride dispersion in dry refluxing ether, the starting material (90%) was recovered. Ethyl-2-pentylp~yridine ( CVH (66). (1) A suspension of sodamide in liquid ammonia (500 mis.) was prepared fran sodium (3.2 g.). To this stirred suspension 5-ethyl-2-methyl pyridine (17 g.) was added over 15 minutes. Stirring was continued for 1 hour and then butyl bromide (19.1 g.) was added as rapidly as possible. The mixture was stirred for 1 hour and the ammonia allowed to evaporate overnight. The residue was taken up in a mixture of ether (200 mis.) and water (200 mls.) and the aqueous layer was made strongly alkaline with 4N sodium hydroxide solution. The ethereal layer was removed and the aqueous layer extracted with ether (4 x 100 mis.). The combined ethereal solutions were dried (K2003), filtered and evaporated. The residue was distilled under reduced pressure. - 117 -

Yield: 8.5 g. (340). b.p. 143°/30 mm. ng3 1.4890. Recovered starting material: 8.0 g. Total base recovery was 75%. (2) A suspension of potassamide in redistilled liquid ammonia (500 mis.) was prepared from potassium (2.8 g.). 5-Ethyl-2-methylpyridine (7.8 g.) and butyl bromide (9.7 g ) wa.oe added as .:bove. Yield: 7 g. (61%). b.p. 143°/30 mm. ng3 1.4891. Recovered starting material 2.5 g. Total base recovered 84%. (Found: 0, 81.2; H, 11.3; N, 8.2. C12H19N requires C, 81.3; H, 10.8; N, 7.9%.)

y 9.00 (T, 0113.0112.0H2), 8.78 (T, 0113.0H2.Ar.) 8.65 (M, (CH2)n), 7.45 (T, m-01-12 .Ar), 7.43 (Q, (3-CH2Ar.)!, 3.11 (D, 3-H), 2.72 (Q, y-H), 1.75 (D, a-11).

This was prepared by method (2) above from potasexm (2.9 g.) in redistilled ammonia (500 mis.) with 5-hex- 5'-eny1-2-methylpyridine (12 g.) and 1-bromopent-4-ene (11.2 g.). Yield: 7.7 g. (47%). b.p. 105°/0.1 mm. ngl 1.50"' (Found: 0, 83.2; lit 10.2; N, 5.9. C17H25N requires

0, 83.8; H, 10.4; N, 5.8%.)

N) max. 1638 (C=C), 1601, 1564, 1495 (Ar.), an.-1. - 118 -

17,7.7-8.7 (id, (0H2)n),, 7.43 (T, (3-0112.Ar.), 7.34 (T, a-CH2.A-2.) ca. 5.1 (14, CH2=), ca. 4.4 (M, 0H=), 3.09 (D, 13-H), 2.73 (Q, Y-H), 1.77 (D, a-H).

5-Dodec-11f-eny1-2-methylpyridine (4 g.), in carbon tetrachloride (50 mis.), was treated at 0° with a solution of bromine (2.0 g.) in carbon tetrachloride (13 mis.) over 15 minutes. The mixture was allowed to stand at 0° for 12 hours and then sufficient chloroform was added to dissolve the oily liquid which had separated. The chloro- form-carbon tetrachloride solution was washed with 10% sodium thio sulphate solution (50 mls.), 1% sodium carbonate solution (10 mls.) and, finally, water (10 mls.). The solution was dried ( T200_), filteredand evaporated. Yield: 6 g. crude (85%). \)max. 1600, 1565, 1495 (Ar.) cm.-1. 8.73 (S,(01-12)n), 7.58 (S, Me.Ar.), 7.50(T, CE2.Ar.), ca. 6.4 (Id, CH2.Dr), ca. 6.1 (M, .0133r.), 3.10 (D, i3-H), 2.79 (Q, Y-H), 1.82 (D, a-H). From the crude product, a picrate was prepared. m.p. (from ethanol) 87.5-89°. (Found: 0, 44.7; H, 5.1; N, 8.6; Br, 25.7.

C24H32Br21\T407 requires 0, 44.5; H, 5.0; N, 8.7; Br, 24.7%.) -119 -

2,5-Di-(5,6-dibromohexv1)Dyridine (LXIX). This was prepared as above from 2,5-dihex-51-anyl pyridine (2.5 g.) and bromine (3.3 g.). Yield: 5.7 g. crude (91%). No satisfactory derivative could be prepared for this tetrabromide. max. 1600, 1563, 1495 (Ar.) cm.-1. No peak at 1638 or 920 cm.-1 . Ts 7.0-8.5 (Us (0112)n), ca. 6.3 (M, CH2.Br), ca. 5.8 (U, Cpr.), 2.95 (Ds 01-H), 2.42 (Qs Y-H), 1.71 (D, a--H). 5-Dodec-111-ynyl-2-methy1pyridine (LXIII). A suspension of sodamide in redistilled liquid ammonia (500 mis.) was prepared from sodium (1.2 g.). Crude 5-(11,12-dibromododecyl)-2-methylpyridine (4.3 g.), dissolved in dry ether (10 mis.) was added to the stirred suspension over 15 minutes. The mixture was stirred for 15 minutes and then solid ammonium chloride (10 g.) was added in portions. The ammonia was allowed to evaporate overnight and the residue taken up in a mixture of ether (100 mis.) and water (100 mis.). The aqueous layer was made alkaline with dilute sodium hydroxide solution and the ethereal layer removed, dried (K2CO3), filtered and evaporated. The residue was distilled under vacuum. -120 -

Yield: 1.1 g. (42%). Overall yield from olefin was 39%. b.p. 124V.05 mm. ng3 1.4952. max. 3350 (CmC-H), 2130 (CmC), 1604, 1565, 1495 (Ar.) cm.-1. , 8.70 (S, (CH2)n), 8.23 (T, CEC-H), ca. 7.9 (111, CH2.0s101.1), 7.56 (S, Me.Ar.), 7.50 (T, CH2.Ar.), 3.11 (D, (3-H), 2.78 (Q, Y-H), 1.81 (D, a-H). Picrate: m.p. (from ethanol) 81°. (Found: N, 1] '5' C24H30N407 requires N, 11.5%.) 2,5-Dihex-5t-YnY1Pyridine (LXX). This was prepared as above from crude 2,5-di-(5,6- dibromohexyl)pyridine (5.7 g.) and sodium (2.2 g.). Yield: 1.1 g. (49%). Overall from diene 45%. b.p. 115°/.04 mm. (Found: 0, 84.7; H, 8.8; N, 5.8. 017H21N requires C, 85.4; H, 8.8; N, 5.8%.) %nax. 3360 (SC-H), 2140 (C;), 1603, 1565, 1495 (Ar.) an.-1. 7E, 7.6-8.6 (M, (CH2)n), 8.25 (T, Ci07.), 7.46 (TO-CE2.Ar.), 7.35 (T, a-0U2.Ar.), 3.04 (D, 8-H), 2.68 (Q,Y-H), 1.72 (D, a-H). 1,24-Di-(2-methyl 5-pyrislyl)tetracosane (LXIV). 5-Dodec-111-yny1-2-methylpyridine (370 mg.) was dissolved in a mixture of ethanol (10 mis.) and concen- trated hydrochloric acid (1 ml.). This solution was added to a mixture of ammonium dhloride (8 g.) and cuprous - 121 -

chloride (5 g.) in water (25 mls.). This mixture was shaken in oxygen until absorption of oxygen ceased (430 mis. at 770 mm. 23°) and then made alkaline with 4N sodium hydroxide solution. The alkaline mixture was filtered, the filter cake washed with chloroform and the filtrate extracted with chloroform (2 x 20 mis.). The combined chloroform solutions were dried (M2003), filtered and evaporated. The residue was dissolved in ethanol (50 mis .) and 2% Pd. on S12003 (147 mg.) was added. The mixture was hydrogenated at atmospheric pressure until uptake of hydrogen ceased and filtered. The filter cake was washed with hot ethanol (100 mls.). The filtrate was evaporated to 25 mis. and cooled. The crystals were filtered. Yield: 120 mg. (32%). m.p. (from 0014) 94-96°. (Found: C, 82.8; H, 11.3; Mol.Wt.(Rast) 521. 036H60N2 requires 0, 82.9; H, 11.6%; Mol.Wt. 509). °max. 1602, 1565, 1495 (Ar.) cm.-1. , 8.74 (5,(0112)n), 7.50 (S, Me.Ar.), 7.46 (T, CH2.Ar.), 2.96 (D, p-H), 2.56 (Q, Y-H), 1.66 (D, m-H). - 122 -

Coupling of 2,5-dihex-5f-ynyllovridine.

Salt diyne Solvent Method T° pH Time Product No. in g. in mg. mis. hrs.

1 CuCl EtOH shaken 25 4 6 no ,7, (70) in coupling O 02 NH4C1 ' ' H2 (10) (1500)

2 CuCi EtOH shaken 35 4 2 as above (180) 223 (70) in NH4C1 H2O 02 (200) (1500)

3 Gun EtOH stirred 55 5 6 as above (40) 520 (140) with NH4C1 H20 02 (80) (3000)

4 OuCi EtOH stirred 25 4 6 as abov (40) P Q (70) with "='' O 0,, NH401 H2 (80) (1500)

5 CuOl EtOH shaken 25 8 6 as abov (5) 250 (70) in NH H20 02 (10) (1500)

6 04.1(0Ac)2 Py. stirred 80 - 72 no 200 (pure) with identi- ?I'Z 50 02 fiable b product -123 -

No. Salt diyne Solvent Method pH Time Product in g. in mg.. mis. hrs.

7 Ou(OA02 316py. vigor- 55 - 3 inter. (41) ous (50%) H2O s±:2-r- S.M. (4.7) ing (50%).

8 Cu(0A0.2 458 Py. stirred 55 - 1 40 inter. H2O (80) vigor- (60%) ously S.M. (9.4) diyne (40%). added 13 mgs./ hr.

9 Cu(0A02 321 py. stirred 55 - 8 as 7. ry (41) vigor- (4.7) ously . . 10 CuCl 350 py. shaken 55 - 8 inter. (3) (50) in 02 (80%) S.M.(20%). ' - , ' ' 11 CuCi 370py. as 10 55 - 30 inter. (3) (1000) (80%) S.M.(20%) . , 12 Cu(0A02 hex- liq. stirred -40 - 3 no dry 1-yne NH3 vigor- coupling (12) 3 g. (200) ously ... , _ L3 Cu(0A02 hex- y. stirred 80 - 12 dodec-5,7- H 0 1_,y,n e pure) with diyne '2 3 g. 85) 0 (60)% (.5) 2 a

Notes: The pyridine was dry commercial material unless otherwise stated. - 124 -

inter = intermolecularly coupled product S.M. = Starting Material (diyne). a _21 1.4894. Lit (98) n19 1.4890. b see reference (99). Work-up of Coupled Product. (1)In all cases where aqueous solvents were used the solution, after the attempted coupling, was made strongly alkaline and extracted with chloroform. The products were analysed by thin layer chromatography, infra-red spectroscopy and, on reduction of the diyne with 2% Pd

on Sr003 in ethanol, by n.m.r. spectroscopy. (2)In all cases, except 6 where the pyridine was evaporated under reduced pressure, where pyridine wasthe solvent) the solution was filtered and the filter cake washed with benzene. The benzene-pyridine mixture was then washed with water until a pyridine-free benzene solution was obtained. The product was analysed as above. When appreciable coupling had occurredl a sample of the product was chromatographed on a thick layer plate (G.F.254 3 mm. developed with CHC13). As far as possible each band was isolated and examined separately. From the reductions of some of the products 2,5- dihexyl pyridine could be isolated. -125-

max. 1600, 1565, 1495 (Ar.) cm.-1. 7:„ 9.1 (T, 0113.0112.), 8.63 (St (C112)n), 7.44 (T, p-CH2.Ar.), 7.34 (T, m-CH2Ar.), 3.10 (D, 0.•.H), 2.74 (Q, Ar-H), 1.77 (D, m-H). Ethyl 6-methylnicotinate (XVIII). (a) From /5-Ethyl-2-methylpyridine (aldehydo (i)The method of Graf. (71). 5-Ethy1-2-methylpyridine (69 g.) was oxidised by a solution of potassium permanganate (360 g.) in water (18 1.) to yield the crude acid hydrochloride (26 g.). This was esterified without purification using ethanol (52 mis.) and concentrated sulphuric acid (23 mis.) (100). Yield: 14 g. (16%, based on 5-ethyl-2-methylpyridine). (ii)The method of Murahashi and Otuka (70). 5-Ethyl-2-methylpyridine (50 g.) was refluxed with concentrated nitric acid (412 mis.) containing ammonium variadate (.2 g.) for three days. The crude acid hydrochloride (31.4 g.) was esterified using ethanol (150 mis.) and dry hydrogen chloride (69). Yield: 20 g. (7n:- 1.- ,Nr,r1 5n 5-ethyl-2-methylpyridine). (iii)The method of Castle and Whittle (69). 5-Ethyl-2-methylpyridine (407 g.) was oxidised by a solution of potassium permanganate (2.087 Kg.) in water (72.5 1.). The mixture of acid and inorganic salts -126 - obtained by evaporation after oxidation was ground to a fine powder and the whole esterified using ethanol (2 1.) and hydrogen chloride. Yield: 204 g. (36.5% based on 5-ethyl-2-methylpyridins). (b) From 5-acetyl-2-methylpyridine (101). This was the method of Bernary and Psille. 5-Acetyl- 2-methylpyridine (82.3 g.) was oxidised by concentrated nitric acid (1470 mis.). The crude acid hydronitrate was esterified using ethanol (220 mls.) and concentrated sulphuric acid (94 mis.) (100). Yield: 42 g. (42% basedon 5-acetyl-2-methylpyridine). In all the above experiments the physical constants of the ester (see below) were similar. The infra-red spectrum of each sample was identical. b.p. 80°/.9 mm., 52-54°/.1 mm., 76-76°/.6 mm. n17 1.5036 n25 1.5021, n190 °5 1.5028. Lit. b.p. 116-118°/11 mm. np0 1.4989. (Found: C, 65.8; H, 6.7; N, 8.5. Cale. for 09 H11 NO 2° • 0, 65.5; H, 6.6; N, 8.5%.) -1 )max. 1710 (0=0), 1600, 1560 (Ar.) an. . 8.61 (T, CH3.CH2.), 7.44 (s, Me-), 5.66 (Q,OH3.01-12-), 2.90 (D, 3-H), 1.95 (Q, Y-H), 1.04 (D, m-H). Baeyer-Williger oxidation of 5-acetyl-2-methylpyridine. 5-Acetyl-2-methylpyridine (8 g.) was dissolved in a mixture of glacial acetic acid (150 mis.) and concentrated -127 -

sulphuric acid (90 mis.). The stirred solution was treated at 20° with a solution of peracetic acid (0.118 moles) in glacial acetic acid (102). The mixture was stirred for 1 hour and then made alkaline with saturated sodium carbonate solution. The alkaline solution was extracted with ether (5 x 200 mis.). The combined ethereal solutions were dried (M004), filtered and evaporated. The residue was distilled under reduced pressure. Recovery of starting material: 4.8 g. (60%). b.p. 118-122°/24 mm. ngl 1.5308. Previous value: no2 2 1.5312. )rnax. 1680 (0=0) am.-1. No absorption at 1710 cm.-1. 6-Metbanicotinamide (LXXI) (57). Ethyl 6-methylnicotinate (19.5 g.), on treatment with concentrated ammonia solution, yielded the amide. Yield: 13.7 g. (85%). m.p. (from water ) 195°. Lit. m.p. 194-196°. 6-Meth lnicotinonitrile (XX) (57). 6-Methylnicotinamide (13.5 g.) was dehydrated by phosphorous oxychloride (14 mis.) to give the nitrile

(XX)• Yield: 10 g. (85%). b.p. 120°/26 mm., m.p. (sublimed) 80°. -128 -

Lit. b.p. 98-99.5°/11 mm. nr.p. 83-84.5°. The dehydration could be conducted in a Carius tube or, on a larger scale, in a lined autoclave. The yields were consistently 85-90%. The purest product was obtained by large scale sublimation at 100° under water-pump vacuum. (Found: Cy 70.8; H, 4.7. Calc. for C7 H6 N2•' C, 71.1; H, 5.1%.) max. 2200 (CmN), 1600, 1490 cm.-1. 1:, 7.28 (S,Me), 2.80 (DO-H), 2.26 (Q,Y-H), 1.28 (D,m-H). Sodioformylacetone. (62). This was prepared by the method of Perez-Medina from sodium (46 g.), acetone (116 g.) and ethyl formate (148 g.). Yield: 177 g. (82%) crude. 3-Cyano-6-methyl-2-pyridone (LXXII) (62). This was prepared from sodium (32 g.), ethyl formate (103 g.), acetone (80.4 g.) and cyanoacetamide (106 g.). Yield: 90 g. (55%). m.p. 286° (decomp.). Lit. 294-296° (decamp.). (Found: C, 62.7; H, 4.5; N, 21.6. Caic. for C7H6N20: C, 62.7; H, 4.5; N, 20.9%.) g-chior nethicunemucx.....1(62). 3-Cyano 6-methyl 2-pyridone (10 g.) was dissolved in phosphorous oxychloride (31 g.) and phosphorous penta- chloride (31 g.) was added. The mixture was heated on

-129 - a steam bath for 1.5 hours and allowed to cool. The excess of phosphorous oxychioride was distilled off under reduced pressure and the residue was poured onto ice (300 g.). The mixture was neutralised with solid sodium bicarbonate and then left at 0° overnight. The solids were then filtered off, dried and continuously extracted with petroleum ether. The solvent was evaporated. Yield: 6 g. (53%). m.p. (from petroleum ether) 111-112°. Lit. m.p. 114-115°. (Pound: C, 54.8; H, 3.4; N, 18.6. Caic. for C7H501N2; C, 55.1; H, 3.3; N, 18.0.) ax. 2220 (CmN), 1588, 1550, 1455 (Ar.) em.-l. N, 7.36 (S, Me), 2.72 (D, ft -H), 2.06 (D, Y-H). 2 -Chloro -3 -cyano -6 -trichloromethlpyridine (Lxxv) (72). 3-Cyano-6-methyl-2-pyridone (87 g.) was treated with phosphorous pentachloride (610 g.) according to the instructions of Perez-Medina for the preparation of 3-cyano-6-methyl-2-chloro pyridine. The product isolated was a tetradhloro compound. Yield: 110 g. (65%). m.p. (from petroleum ether) 80-85°. (Found: 0, 33.5; H, 1.3; Cl, 54.7. C7H2C14N2 requires C, 32.9; H, 0.8; Cl. 55.5%.) max. 2220 (C-17), 1585, 1545, 1440 (Ar.) cm.-1. TN, 1.76 (D, t3-H), 1.94 (D, Y-H). -130 -

Attempted selective reduction of 2-chloro-3-cyano-6-methyl- pyridine (103). (1) 2-Chloro-3-cyano-6-methylpyridine (9 g.) was hydro- genated at 3.5 atmospheres and 25°C using freshly prepared palladium chloride (138 mg.). This method was that of Bobbit and Scola. The starting material was recovered (670). (2) 2-Chloro-3-cyano--6-methylpyridine (5 g.) was hydro- genated at atmospheric pressure using 1% Pd on CaCO3 (5 g.) in ethanol (50 mis.). Again starting material was recovered. (3) 2-Chloro-3-cyano-6-methylpyridine (10.5 g.) in dry ether (150 mis.) was treated with a solution of lithium aluminium hydride in dry ether (24 mis. containing 0.755 moles/i.) without cooling. Gentle refluxing occurred and when the addition was complete the excess of hydride was decomposed with water (10 mis.). The ethereal layer was removed and the aqueous layer extracted with chloroform (2 x 50 mis.). The combined organic solutions were dried (MgSO4), filtered and evaporated. Starting material was recovered(60'0), In all the methods used above, therewas no evidence of any 6-methylnicotinonitrile in the products. -131 -

2-Chloro-3-11 -ozaR20y1-6-m2jhapyridine (XXIV) (107). 2-Chloro-3-cyano-6-methylpyridine (3.2 g.) was treated with butyl magnesium bromide, prepared from butyl bromide (5.64 g.) and magnesium(1.0 g.). The method was that of Mariella and Kvinge. A low yield of a steam -1 volatile oil was obtained. It had max. 1682 am. and no peak at 2220 cm.-1 (O=N) or 1660 (pyridone) 5-Acetyl-2-methylpyridine (LXXVI) (101) This was prepared by treating sodioformyl acetone (177 g.) with a saturated solution of ammonium acetate (98 g.) :.1.a glacial acetic acid according to the method of Bernary and Psille. Yield: 26 g. (23%). b.p. 227-2310/760 mm., 118°/28 mm. 2 no 1.5312. 20 Lit. b.p. 231-232°/760 mm., 108-110°/10 mm. no 1.5308. (Pound: C, 70.7; H, 6.7; N, 10.2. Calc. for C8H9NO:

Cy 71.1; H, 6.7; N, 10.3%). \) max. 1685 (C=0), 1599, 1562, 1495 (Ar.) cm. -1 , 7.43 (S,CH3C0), 7.40 (Sp Me.Ar.), 2.80 (D,3-H), 1.93 (Q, Y-H), 0.99 (D, m-H). chlorpmethyl ethyl ether (104). This was prepared by treating a mixture of dry ethanol (69 g.) and paraformaldehyde (45 g.) with dry -152 -

hydrogen chloride at 0°. Yield: 105 g. (67%). b.p. (under hydrogen) 80-84°/765 mm. n22 1.0259. Lit. b.p. 81-82°/760 mm. n12 1.0282. 2-ChloroetIva ethyl ether (XXIX) (105). This was prepared by treating 2-ethoxyethanol (717 g. in pyridine (615 g.) with thionyl chloride (650 mis.) at 0- -5°. Yield: 550 g. (63%). b.p. 108-109°/760 mm. n2° 1.4112. Lit. b.p. 105-110°/760 mm. n2° 1.4113. 194-Dibromobutane (106.) (a) A mixture of peroxide-free tetrahydrofuran (555 g.), water (120 mis.) and red phosphorous (92.6 g.) was treated with bromine (352 mis.). Yield: 1210 g. (80%). b.p. 102-103°/38 mm. Lit. b.p. 83-84°/12 mm. (b)A mixture of butan-1,4-diol (750 g.) and red phos- phorous (114 g.) was treated with bromine (415 mis.) at 110°. Yield: 1140 g. (63%). Physical constants were as above. 4-21romo-ethoybutane (pc7) n=21 (108,109). 1,4-Dibromobutane (1140 g.) on dry ethanol (500 mis.) was treated, under reflux with a solution of sodium methoxide in dry ethanol, prepared from sodium (107 g.) - 133 -

and dry ethanol (1800 mls.). Yield: 270 g. (32%). b.p. 60-64°/76 mm. n23 1.4482. Lit. b.p. 169°/760 mm. nn0 1.4490. max. 1110 (ether) cm.-1. izahampentan-l-ol (110 111). 4-Ethoxybutyl magnesium bromide, from 1-bromo-4- ethoxy butane (110 g.) and magnesium (16.2 g.), in dibutyl ether was treated with paraformaldehyde (60 g.) at 100-110°. 22 Yield: 9 g. (10%). b.p. 44-47°/2 mm. n 1.4289. Lit. 94-97°/14 mm. ng° 1.4291. '\) max. 1110 (ether) am.-1. 5-Chloro-1-7ethoxypentane (112, 105). This was prepared by treating 5-ethoxypentan-l-ol (9.0 g.) in pyridine (5.4 g.) with thionyl chloride (8.9 g.). Yield: 6 g. (60c:'). b.p. 60-64°/8.5 mm. -Alo22 1.4290. 20 Lit. b.p. 72°/17 mm. no 1.4300. -1 max. 1110 (ether) cm . . 1-EthoxViDent-4-ene (LnqX; n=1), (113,114). This was prepared by treating but-3-enyl magnesium bromide, prepared from 1-bromobut-3-ene (105 g.) and magnesium (18.7 g.) in dry ether (400 mis.), with chloromethyl ethyl ether (72 g.) in dry ether at 0°. -134- The method is that of Waterman et al. Yield: 56.1 g. (65%). b.p. 116-120°/160 mm. ng4 1.4052. Lit. b.p. 117-118 °/760 mm. ng0 1.4082. max. 1639 (0=0), 1110 (ether) cm.-1. -Bromo-l-ethoxypentane (LXXX§ n=31 (115,109). 1-Ethoxypent-4-ene (53.1 g.) was dissolved in dry benzene (600 mis.), containing a trace of dibenzoyl peroxide. The solution was kept at 4o while a rapid stream of bromine-free hydrogen bromide was passed in. When the solution was saturated with hydrogen bromide, the passage of gas was stopped and the solution washed with water (100 mls.), 5% aqueous sodium carbonate solution (150 mis.) and water (100 mis.). The benzene solution was evaporated at atmospheric pressure and the residue distilled under reduced pressure. Yield: 65 g. (64%). b.p. 800/23 mm. n25 1.4528. Lit. b.p. 85°/14 mm. n30 1.4520. max. 1110 (ether) cm.-1. 1-Ethoxyhect-6,-ene_ (LXXXIulaal. This was prepared by treating 4-ethoxybutyl magnesium bromide, prepared from 4-bromo-l-ethoxybutane (90 g.) and magnesium (11.6 g.), in dry ether (400 mis.) with allyl bromide (250 mis.) in dry ether (250 mis.). The method was that used for the hydrobromination of - 135 - tetradeca-1,13-diene (LIX). Yield: 45 g. (65%). b.p. 64°/18 mm. n1 1.4201. max. 1640 (C=C), 1110 (ether) cm.-1.

1-, 8.88 (T, CH3.0112.0), 8.64 (SP(CH2)n), 6069 (T,011200H2.0.) 6.64 (Q,CH3.0112.0.), ca. 5.1 (M,CH2=), ca. 4.4 (M,CH2=CH). 7-Bromo-l-ethoxyheptane (LXXXi n=5). 1-Ethoxyhept-6-ene (45 g.) was treated with hydrogen bromide in benzene at 4° as in the preparation of 11- bromo-1-ethoxyundecane (LXXX; n=9). Yield:_ 58 g. (82%). b.p. 51°/.13 mm. n22 1.4548. (Found: Br, 36.8. 09H190Br requires Br, 35.9%). 17, 8.87 (T, CH3.CH2), 8.60 (S,(CH2)n), 6.69 (T,CH2CH2-0)

6.68 (T, CH2Br), 6.65 (Q, CH30H2--0). ---2But--e11- xxvil (74,75). A mixture of magnesium (48 g.) and paraformaldehyde (60 g.) in dry ether (550 mis.) was treated under reflux with ailyl bromide (240 g.). Yield: 47.4 g. (30%). b.p. 114-116°/760 mm. n1 9 1.4211. Lit. b.p. 112-114°/760 mm. Pent-4-en-l-ol (116,117,118). (a) Allyl magnesium chloride, prepared from magnesium (80.5 g.) and allyl chloride (231 g.), in dry ether (800 mis.) was treated with ethylene oxide (154 g.) at 0°. Yield: 60 g. (24%). b.p. 140-141°/760 mm. ngl 1.4291 - 136 -

Lit. b.p. 134-137°/751 mm. nr 1.4299. (b) (i) Tetrahydrofurfuryl chloride was prepared by treating tetrahydrofurfuryl alcohol (454 g.) in dry pyridine (398 mis.) with thionyl chloride (560 g.) at 0°. Yield: 252 g. (420). b.p. 54°/20 mm. Lit. b.p. 47-48°/15 mm. (ii) Tetrahydrofurfuryl chloride (190 g.) was added to sodium sand (71 g.) covered with dry ether (450 mis.) to yield pent-4-en-l-ol. Yield: 100 g. (730). b.p. 135-140°/760 mm. n22 = 1.4308. Lit. b.p. 76.4°/60 mm. ng° 1.4299. max. 1638 (C=0) au.-1.

1-Bromobut-3-ene (LXXVIII3 n=1) (74,119). This was prepared by treating but-3-en-l-ol (86.4 g.) in dry pyridine (26.4 g.) with phosphorous tribromide (129 g.) at 20°. Yield: 110 g. (77%). b.p. 99-101°/769 mm. n23 1.4603. Lit. b.p. 97-102°/760 mm. nn7 = 1.465. 2...... Bromoen acvnin=2j (118). This was prepared by treating pent-4-en-1-ol (56 g.) in pyridine (14 g.) with phosphorous tribromide (70.6 g.) at 0°. Yield: 54 g. (57%). b.p. 123-129°/760 mm. n23 1.4598. Lit. b.p. 125-129°/760 mm. n20 1.4632. max. 1638 (0=0) cm.-1. - 137 -

1-Chloro ent-4-ene (75). This was prepared by treating pent-4-en-l-ol 00 g. in bromobenzene (126 mls.) containing pyridine (1 ml.) with thionyl chloride (136 g.).

1 . 42,91 . Yield: 82 g. (62%). b.p. 104-106.5°/160 nod Lit. b.p. 104-105°/760 mm. n18.5 1.4309. -1 max. 1638 (c=c) aa. . - 138 -

The 1,4 Bridged Pyridine

Experimental Section -139-

1.-Dodec-1l'i..... -.eny22..WItp_1,e(LXXXII (76). A suspension of sodamide in liquid ammonia (400 mls.) was prepared from sodium (83 g.) (120). 4 -Picoline (28.5 g.) in dry ether (25 mis.) was added to the stirred suspension over 15 minutes and the mixture stirred for 30 minutes. 1 -Chloroundec -10 -ene (68 g.) in dry ether (75 mis.) was added over 15 minutes. The mixture was stirred for 45 minutes and the ammonia allowed to evaporate overnight. The residue was taken up in ether (100 mls.) and water (100 mis.). The ethereal layer was removed and aqueous layer extracted with ether (3 x 50 mis.).

The combined ethereal solutions were dried (E.2CO3) filter--.d and evaporated. The residue was taken up in petroleum ether (200 mls.) and dry hydrogen chloride passed into the solution. The precipitate was filtered off, washed with ether (50 mls.) and recrystallised from acetone. Yield: 23.6 g. (27%). m.p. (from acetone) = 129-130°. (Found: 0, 72.7; H, 9.9. 0171128N.01 requires C, 72.4; 11, 10.0%.) -1 max. = 1638 (Ar. and 0=0), 1604, 1505 (Ar.) cm. The free base was isolated by treatment of a pure sample of the hydrochloride with dilute sodium hydroxide solution and extraction with ether. The ethereal - 140 -

solution was evaporated and the residue dried over phos- phorous pentoxide under vacuum. n22 = 1.4908. Lit. n14 = 1.4940.

(Found: C, 83.5; H, 10.9; N, 5.7. Cale. for 017H27N: C, 83.2; H, 11.; N, 5.7%.) -1E, 8.73 (s, (OH2)12), 7.44 (T, Y-0H2.Ar.), ca. 5.1 (M,CH2=), ca. 4.4 (11, 011=)„ 2.96 (D, P-H), 1.51 (D, a-H). 4-11,-Bromododecylpyridine (LXXXIII). 4-Dodec-111 -enyl pyridinium hydrochloride (15 g.) was dissolved in 48% aqueous hydrobromic acid (150 mis.). The solution was allowed to stand at 200 for 40 hours and then made alkaline with solid sodium carbonate. The mixture was extracted with benzene (1 x 150, 4 x 50 mis.). The combined benzene solutions were filtered and evapor- ated to 150 mis. at atmospheric pressure. The cooled solution was passed through an alumina column (20 am. x 2 cm. Type H). The column was eluted with benzene. The first 300 mis. of eluate were evaporated under :.3duceC, pressure. The residue was a colourless oil. Yield: 11 g. (63%). n16 = 5111. (Found: C, 63.2; H, 8.8; N, 4.1. C17H28N.Br requires 0, 62.5; H, 8.6; N, 4.3%.) cm. -1 max. = 1603, 1560, 1500 (Ar.) T P 8.70 (S,(0H2)n), 8.32 (D, 0',.:3.CHBr.), 7.43 (T, Y-0112.) 6.00 (U,01-13.0HBr.), 3.00 (D, p-H), 1.59 (D, a-H). - 141 -

12,27-dimethvltric,clo(26,2,2,213,16.)_ 13,28-diazocinium tetratricontane-13 15 28 30 31 34-hexaene dibromide (LXXXIV). 4-111-Bromododecyl pyridine (4.9 g.) was heated in a high vacuum retort at 150° under .03 mm. After 15 minutes the temperature was raised slowly until the liquid began to spit. The apparatus was allowed to cool and the product scraped from the walls. It was recrystallised from carbon tetrachloride. Yield: 2 g. (41%). m.p. (from carbon tetrachloride)= 124-125°.

(Found: 0, 62.6; H, 8.7; N, 4.3; Mol.Wt.(benzene), 669. (C17H28NBr)2 requires C, 62.5; 119 8.6; N, 4.3; Mol.Wt., 652.) max. = 1633, 1600, 1504 (Ar.) cm.-1.

9.05 (DI, ?), 8.89 (S, ?), 8.70 (S,(CH2)n), 7.03 (T, y-0112), 4.60 (U, CH.N), 2.13 (D, 3-H), 1.08 (D, a-H). 12,27 imethyltricyclo(26 2,2,2 13'16 . _ 13,28-diazo- tetratricontane (LXXXVI). The cyclic quaternary salt (LXXXIV) (940 mgs.) was dissolved in water (20 mis.) and Adam's catalyst (35 mgs.) was added. The mixture was hydrogenated at atmospheric pressure. The solution was made alkaline with 2N sodium - 142 -

hydroxide and extracted with chloroform (3 x 20 mls.).

The combined chloroform solutions were dried (Na2SO4), filtered and evaporated. The rosidue was sublimed at 100° under .02 mm. Yield: 400 mgs. (550). m.p. 59-60°.

(Found: 0, 79.6; H, 13.3; N, 5.7; Mol.Wt.(Rast), 467. C34H 66 N21-2112 0 requires C, 79.7(5); H, 13.2; N, 5.50; Mol.Wt. 5.11). 17, 9.10 (D, CH3.CH), 8.73 (Sp(0112)n), 6.6-7.5 (IT, CE2.Np cE.N). 1-Bromo-11-chloroundecane (LXXXVIII). 1-Chloroundec-10-ene (50 g.) in benzene (350 mis.) was treated with hydrogen bromide at 4° as for 1-ethoxy- undec-10-ene. Yield: 52 g. (73%). b.p. 108-113°/.35 mm. n18 1.4772. (Found: Cy 49.2; H, 8.4.C11H22 BrC1 requires Cp 49.0; H, 8.20). T , 8.70 (Sp(CH2)n), 6.65 (T, CH2Br), 6.52 (T, CH2C1). 4-121-Chlorododecyluridine (XI; x=C1). This was preparedp by the method used for 4-dodec- 11t-enyl pyridine, from 1-bromo-11-chloroundecane (50.8 g.) 4-picoline (15.3 g.) and sodium (4.3 g.) in liquid ammonia (500 mls.). Yield: (as hydrochloride) 43.5 g. (840). - 1.43 -

The free base was a colourless liquid which was distilled on a short path at 140°/.07 mm. ngl 1.4984. (Found: N, 4.8. C17H28C1N requires N, 4.9%.) \) max. 1602, 1560, 1495 (Ar.) cm.-1. 1, 8.73 (8,(CH2)n), 7.42 (T, Y-CH2), 6.56 (T, CH2C1), 3.00 (D, 3-H), 1.59 (D, m-H). Attem ted intramolecular oyclisation of 4-121-dhloro- dodecylpyridine. A solution of the free base in xylene (150 mis.) was refluxed for 24 hours. The solvent was removed under vacuum. Starting material was recovered in good yield. The same result was obtained using p-cymene as solvent. When 1-methyl naphthalene was used, charring occurred and no identifiable material was isolated. l-Iodoundec-l0-ene (85). This was prepared from 1-chloroundec-10-ene (90.5 g.) by treatment with sodium iodide (141 g.) in dry acetone (850 mis.), according to the method of Brody and Bogert. Yield: 88.2 g. (67%). b.p. 93-97V 5 mm. no23 1.4931. Lit. b.p. 104°/2 mm. n(2)5 1.4937. 1-ataX2Pdec-10-e/121IZZIKL2a7i- Sodium (6.9 g.) was dissolved in dry ethanol (130 g.). To this solution 1-iodoundec-10-ene (84 g.) was added -144-

with stirring. When the addition was complete, the mixture was refluxed for two hours, cooled and diluted with water (250 mis.). The aqueous solution was extracted with n-hexane (1 x 100 mis., 3 x 50 mis.). The n-hexane solution was dried (0a012), filtered and evaporated. The residue was distilled under vacuum. Yield: 39 g. (66%). b.p. 86°/.9 mm. ng3 1.4324. Lit. b.p. 114-11574 mm. 4)1 1.4324. (Found: CI 78.9; H, 13.2. Calc. for C13H260: C, 78.7;

H, 13.2). )max. 1639 (0=0), 1110 (ether) cm.-1. T, 8.87 (T, CH3.CH2.0), 8.70 (S1(CH2)n), 6.69 (T,CH2.0112.0.), 6.64 (Q, CH3.0112.0.), ca. 5.1 (M, CH2=), ca. 4.4 (M, CH=). 11-Bromo-l-ethocyundecane (LXXXt_ n=9). 1-Ethoxyundec-10-ene (35.5 g.) was dissolved in dry benzene (300 mis.) containing a trace of dibenzoyl peroxide. Into this solution, kept at 4°C., a rapid stream of bromine-free hydrogen bromide was passed until the gas was no longer absorbed. The benzene was evapor- ated under reduced pressure and the residue was distilled under vacuum. Yield: 46 g. (92%). b.p. 106-110°/.15mm. nil_0 1.4591.

(Found: Br, 28.8(5). C13 H27BrO requires Br, 28.6). max. 1115 (ether) cm.-1. -145-

11-Bromoundecyl-p-toluene suiphonate. 11-Bromoundecan-l-ol (5 g.) was dissolved in dry pyridine (10 mis.). This solution, kept at 10°, was treated with p-toluene sulphonyl chloride (4.2 g.), added in portions over 15 minutes. The mixture was stirred at 20° for two hours and then poured into a mixture of con- centrated hydrochloric acid (18.5 mis.) and ice (60 g.). An oil separated and this readily crystallised. Yield: 6.0 g. (75%). m.p. (from petroleum ether) 31-31.5° . (Found: 0, 53.1; H, 7.5. C18H29Br03S requires 0, 53.3; H, 7.2%.) 4-12'-Ethoxydodecylpyridine (LXXXVIII). This was prepared by the same method as 4-dodec-111- enyl pyridine from sodium (1.4 g.), 4-picoline (5.7 g.) and 11-bromo 1-ethoxy undecane (16.7 g.). The product was isolated as the hydrochloride. Yield: 11.8 g. (60%). m.p. (from acetone) 124-126°. (Poland: CI 69.6; H, 10.4; N, 4.5; Cl, 11.5.

019H34C1N0 requires C, 69.7; H, 10.5; N, 4.3; Cl, 10.9.) max. 1640, 1615, 1518 (Ar.), 1110 (ether) cm.-1.

E, 8.82 (T,CH3.CH2.0.), 8.72 (S,(CH2)n), 7.09 (T,Y-CH2),

6.58 (T, CH2.CH2.0.), 6.53 (Q, CH3.0112.0), 2.23 (D, c3—H), 1.20 (D, m-H). -146 -

4-12I-Bromododecylpyridinium hydrobromide (XI, x-Br). 4-(12-Ethoxydodecyl-}pyridinium hydrochloride (1.5 g.) was suspended in 48% aqueous hydrobromic acid and the mixture sealed in a Carius tube. The tube was kept at 130° for 3 hours and then allowed to cool. The contents of the tube were filtered and the crystals washed with ice-cold acetone. Yield: 1.1 g. (60%). m.p. (from acetone) 136-136°. (Found: Br, 39.4. C17H29Br2N requires Br, 39.2.) V ms.r. 1638, 1605, 1520, 1500 (Ar.) cm.-1.

rE, 8.72 (S,(CH2)n), 7.08 (T, y-0112), 6.62 (T, CH2.Br) 2.14 (D, 3-H), 1.05 (D, m-H). An attempt was made to prepare the above compound by treating a solution of 4-picoloyl sodium ±n liquid ammonia (from sodium (.29 g.) and 4-picoline (1.05 g.)) with 11-bromo undecyl p-toluene sulphonate (5 g.). The only pyridine isolated was recovered 4-picoline. 14,30-Diazocinium tricyc12/28)2,2,214,17)hexatricontane- 14,16135,30,32,33-hexaene dibromide (LXXXIX). 4-12'-Bromododecylpyridine (400 mg.) was heated in a high vacuum retort at 150° under .03 mm. The white crystals which formed on the walls of the retort were scraped off and recrystallised from carbon tetrachloride. -147-

Yield: 180 mg. m.p. (from 0014) 110°.

(Found: N, 4.3. (017H28NBr)2 requires N, 4.3%). \)max. 1633, 1600, 1504 (Ar.) cm.-1. TH, 9.13 (s,?), 9.05 (S,?), 8.95 (S,?), 8.72 (S,(CH2.)2, 7.06 (T, Y-0112.), 4.63 (U, CH2.N), 2.02 (D, 0.41), 0.95 (D, a-H). - 148 -

REFERENCES

1. A. Kekule, Annalen, 1872, 162, 77. 2. J. Thomsen, Ber., 1880, 13, 1808. 3. E. Heckel, Z.Physik, 1932, 76, 628. 4. L. Pauling and G. W. Wheland, J.Chem.Phys., 1933, 1, 362. 5. G. W. Wheland, "Resonance in Organic Chemistry", John Wiley and Sons, New York, 1955. 6. C. A. Coulson, "Valence", Oxford Untv.Press, Oxford, 1952. 7. D. P. Craig, 'Non-benzenoid Aromatic Compounds", edited by D. Ginsberg, Interscience, New York, 1959. 8. F. Klarges, Chem.Ber., 1949, 82, 358. 9. H.D.S. Springall, T. R. White, and R. C. Cass, Trans .Faraday Soc., 1954, 50, 815. 10. L. Pauling, L. O. Brockway, and J. Y. Beach, J.Amer. Chem.Soc., 1935, a, 2705. 11. W. G. Penney, Proc.Roy.Soc., 1937, A158, 306. 12. C. A. Coulson, Proc.Roy.Soc., 1939, A169, 413. 13. C. A. Coulson and A. Golebiewski, Tetrahedron, 1960, 11, 125. 14. F. Sondheimer et al., J.Amer.Ohem.Soc., 1962, 4307. -149-

15. E. Mickel, Z.Phyzik, 1931, 70, 204. 16. E. Legoff and R. B. LaCount, J.Amer.Chem.Soc., 1963, E2, 1354. 17. R. A. Raphael, Proc.Chem.Soc., 1962, 67. 18. D. P. Craig, J.Chem.Soc., 1951, 3175. 19. L. M. Jackman, "Applications of N.M.R. Spectroscopy in Organic Chemistry", Pergamon Press, London, 1959. 20. J. A. Pople, W. G. Schneider, and H. J. Bernstein, "High Resolution N.M.R.", McGraw-Hill, New York, 1959. 21. J. D. Roberts, "Nuclear Magnetic Resonance", McGraw-Hill, New York, 1959. 22. P. Pascal, Ann.Chim.Phys., 1910, lap 5; 1912, 25, 289; 1913, 22, 218. 23. L. Pauling, J.Chem.Phys., 1936, 4, 673. 24. J.A. Pople, J.Chem.Phys., 1956, al 1111. 25. J. S. Waugh and R. W. Fessenden, J.Amer.Chem.Soc., 1957, MI 846. 26. C. E. Johnson and F. A. Bovey, J.Chem.Phys., 1958, .2.1, 1012. 27. J. A. Pople, Mol.Phys., 1958, 1, 175. 28. R. McWeeny, Mol.Phys., 1958, 1, 311. 29. J. A. Elvidge and L. M. Jackman, J.Chem.Soc., 1961, 859. -150 -

30. J. A. Elvidge, Chem.Comm., 1965, 160. 31. R. J. Abraham, R. C. Sheppard, W. A. Thomas and S. Turner, Chem.Comm., 1965, 43. 32. V. Prelog et al., Helv.Chim.Acta, 1947, 30, 1741; M.Stoll et al., Helv.Chim.Acta, 1947, 30, 1815, 1822, 1837. 33. D. J. Cram and H. U. Daeniker, J.Amer.Chem.Soc., 1954, 76, 2745. 34. Ya. L. Goldfarb, S.Z. Taits, and L. I. Belen'kii, Izvest.Akad.Nauk S.S.S.R., Otdel.Khim.Nauk, 1957, 10, 1262. 35. M. Nakazaki and K. Yamamoto, Chem.and Ind., 1965, 468. 36. L. Ruzicka et al., Helv.Chim.Acta, 1926, 2, 389, 499; 1928,,11, 496, 670; 1932, 15, 1459. M. Stoll and A. Rouve, Helv.Chim.Acta, 1944, 27) 1570. 37. G. M. Bennet, Trans.Faraday Soc., 1941, 27, 794. 38. P. Ruggli, Annalen, 1912, 222, 92. 39. M. Stoll and A. Rouve, Helv.Chim.Acta, 1947, 30, 2019. 40. B.C.L. Weedon, Quart.Rev., 1952, 6, 380. 41. H. Hunsdiecker, Ber., 1942, 75, 1190, 1197. 42. Ya. L. Goldfarb, S. Z. Taits, and V. N. Balgakova, Izvest.Akad.Nauk S.S.S.R., Otdel.Khim.Nauk, 1963, 7, 1299. -151-

43. W. M. Schubert and W. A. Sweeny, J.Amer.Chem.Soc., 1955, 77, 4172. 44. A. T. Blomquist et al., J.Amer.Chem.Soc., 1947, 69, 472; 1948, 70, 30, 34. 45. K. Ziegler et al., Annalen, 1933, 504, 94; 1934, 512, 164; 1934, 513, 14; 1937, 528, 114. 46. G. Eglinton and A. R. Galbraith, Chem. and Ind., 1956, 737. J.Chem.Soc., 1959, 889. 47. F. Sondheimer, R. Wolovsky, and Y. Amiel, J.Amer. Chem.Soc., 1962, a, 260, 270, 274. See ref.17. 48. Brit.Pat. 938,697/1963. 49. Brit.Pat. i22,409/1963. 50. J. Dale and A. J. Hubert, J.Chem.Soc., 1963, 86. 51. G. Eglinton and W. McCrae, "Advances in Organic Chemistry", Interscience, New York, 1963, V01.4, p.225. 52. Y. Iwakura and K. Una, J.Org.Chem., 1960, 25, 1227. 53. A. Idttringhaus, Annalen, 1937, 528, 181, 211, 223. 54. E. V. Brown, "Pyridine and Its Derivatives", edited by E. Klingsberg, Interscience, New York, 1964, Ch.XIII, p.6. 55. R. S. Tipson, J.Org.Chem., 1944, 235. 56. C. D. Nenitzescu, S. Tzitzeica, and V. Than, Bull.Soc.chim.Fraace, 1955, 1272. -152 -

57. P. A. Plattner, W. Keller, and A. Boller, Helv.Chim. Acta, 1954, 37, 1379. 58. J. P. Wibaut and E. C. Kooijman, Rec.Tray.chim., 1946, 65, 141; 1947, 66, 705. 59. C. Osuch and R. Levine, J.Amer.Chem.Soc., 1956, 78, 1725. 60. M. S. Kharasch and O. Reinmuth, ”Grignard Reactions of Non-metallic Substancesm, Constable and Co., London, 1954, p.36. 61. L. A. Walter, W. H. Hunt, and R. J. Fosbinder, J.Amer.Chem.Soc., 1941, §2, 2771. 62. R. P. Mariella, J.Amer.Chem.Soc., 1947, 69, 2670. 63. D. A. Rosenfield, N. K. Ritchmyer, and C. S. Hudson, J.Amer.Chem.Soc., 1948, 70, 2201. 64. E. P. Pratt and J. D. Draper, J.Amer.Chem.Soc., 1949, 71, 2846. 65. G.I.T., Fachzeitschrift fUr das Laboratorium, 1962, 514. 66. A. D. Miller and R. Levine, J.Org.Chem., 1957, 22, 168. 67. A. A. Clifford and W. A. Waters, J.Chem.Soc., 1963, 3056. 68. G. E. Coates and C. Parkin, J.Inorg.Nuclear Chem., 1961, 22, 59. 69. R. H.Castle and C. W. Whittle, J.Org.Chem., 1959, 1191. 70. S. Murahashi and S. Otuka, Mem.Inst.Sci.Ind.Res., Osaka Univ., 1950, 7, 127. 7?.. R. Graf, J.prakt.Chem., 1932, 10., 19. 72. L. A. Perez-Medina, R. P. Mariella, and S. P. McElvain, J.Amer.Chem.Soc., 1947, a, 2576. 73. S. L. Friess and A. H. Soloway, J.Amer.Chem.Soc., 1951, 73, 3968. 74. R. P. Linstead and H. N. Rydon, J.Chem.Soc., 1934, 1998. 75. M. F. Ansell and S. S. Brown, J.Chem.Soc., 1957, 1788. 76. C. Barkovsky, Ann.Chim.(France), 1944, 22: 487. 77. A. Grtn and T. Wirth, Ber., 1922, .2.2, 2208. 78. F. R. Mayo and C. Walling, Chem.Rev., 1940, 27, 351. 79. R. Ashton and J. C. Smith, J.Chem.Soc., 1934, 435. 80. F. Brady and M. T. Bogert, J.Amer.Chem.Soc., 1943, 65, 1078. 81. T. M. Patrick, J.Amer.Chem.Soc., 1952, ZA, 2984. 82. K. Bieman, G. BUchil and B. H. Walker, J.Amer.Chem. Soc., 1957, 22,, 5558. 83. L. Ruzicka and M. Stoll, Helv.Chim.Acta, 1934, 17, 1308. 84. N. Sperber, D. Papa, E. Schwenk, and M. Sherlock, J.Amer.Chem.Soc., 1951, 22, 3856.

-154-

85. F. Brody and M. T. Bogert, J.Amer.Chem.Soc., 1943, 65, 1080. 86. C. Prrost, Compt.rend., 1933, 196, 1129. 87. V. Prelog, M. M. Wirth, and B. Schbnbaum, Helv.Chim. Acta, 1946, 29, 1204. 88. R. Kapp and A. Knoll, J.Amer.Chem.Soc., 1943, 65, 2062. 89. T. D. Perrine, J.Org.Chem., 1953, 18, 1356. 90. P. Chuit and J. Hausser, Helv.Chim.Acta, 1929, 12, 475. 91. N. Polgar and R. Robinson, J.Chem.Soc., 1945, 389. 92. J. Cason, Chem.Rev., 1947, .4Q, 15. 93. F. Asinger and H. Eckoldt, Ber., 1943, 76, 581. 94. C.S. Marvel and W. E. Garrison, J.Amer.Chem.Soc., 1959, 81, 4737. 95. P. Chuit, Helv.Chim.Acta, 1926, 9, 272. 96. A. N. Nesmeyanov, L. I. Zakharkin, and T. A. Kost, Izvest.Akad.Nauk S.S.S.R., Otdel.Khim.Nauk, 1955, 657. 97. S. Bergstr5m et al., Acta Chem.Scared., 1952, 6, 1170. 98. H. K. Black and B.C.L. Weedon, J.Chem.Soc., 1953, 1785. 99. E. SpAth, Monatsh., 1912, 33, 235. 100. F. B. LaForge, J.Amer.Chem.Soc. 1928, 50, 2477. 101. E. Benary and H. Psille, Ber., 1924, 57, 832. -155-

102. F. P. Greenspan, J.Amer.Chem.Soc., 1947, 68, 907. 103. J. M. Bobbit and D. A. Scola, J.Org.Chem., 1960, 25, 560. 104. J. W. Farren and H. R. Fife, J.Amer.Chem.Soc., 1925, AI, 2419. 105. D. E. Ames and R. E. Bowman, J.Chem.Soc., 1950, 406. 106. A. I. Vogel, "Practical Organic Chemistry", 3rd edition, Longmans, London, 1956, p.283. 107. C. Rath and F. Schiffmann, Annalen, 1931, An, 127. 108. J. T. Grayson, K. W. Greenlee, J. M. Derfer, and C. E. Boord, J.Org.Chem., 1955, 20, 275. 109. S. Oae, J.Amer.Chem.Soc., 1956, 78, 4030. 110. R. I. Longley, W. S. Emerson and T. C. Shafer, J.Amer.Chem.Soc., 1952, 74, 2012. 111. A. I. Vogel, "Practical Organic Chemistry", 3rd edition, Longmans, London, 1956, p.253. 112. P. M. Frearson and E. S. Stern, J.Chem.Soc., 1958, 3062. 113. L. I. Zakharkin and L. A. Savina, Izvest.Akad.Nauk S.S.S.R., Otdel.Khim.Nauk, 1960, 1039. 114. H. I. Waterman, W.J.C. de Kok, J. J. Leendertse, and W. H. Schoenmaker, Rec.Tray.chim., 1937, 56, 437. -156-

115. V. Prelog and R. Seiwerth, Ber., 1939, 72, 1638. 116. M. S. Kharasch and C. F. Fuchs, J.Org.Chem., 1944, 359. 117. L. A. Brooks and H. R. Snyder, 'Organic Syntheses", John Wiley and Sons, New York, 1945, Vol.25, p.84. 118. P. Gaubert, R. P. Linstead, and H. N. Rydon, J.Chem.Soc., 1937, 1971. 119. H. Pariselle, Ann.Chim.(France), 1911, al 322. 120. R. P. Linstead, J. A. Elvidge, and M. Whalley, "Modern Techniques in Organic Chemistry", Butterworths, London, 1955, p.104.