UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Some INDOR and preparative studies of organometallic and carbohydrate derivatives Steiner, Paul Robert 1971

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

Item Metadata


831-UBC_1971_A1 S84_5.pdf [ 6.5MB ]
JSON: 831-1.0059846.json
JSON-LD: 831-1.0059846-ld.json
RDF/XML (Pretty): 831-1.0059846-rdf.xml
RDF/JSON: 831-1.0059846-rdf.json
Turtle: 831-1.0059846-turtle.txt
N-Triples: 831-1.0059846-rdf-ntriples.txt
Original Record: 831-1.0059846-source.json
Full Text

Full Text

SOME INDOR AND PREPARATIVE STUDIES OF ORGANOMETALLIC AND CARBOHYDRATE DERIVATIVES. by PAUL ROBERT STEINER B.Sc, University of British Columbia, 1966 M.Sc. University of British Columbia, 1969 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF .THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in the Department of CHEMISTRY We accept this thesis as conforming to the requir standard THE UNIVERSITY OF BRITISH COLUMBIA February. 1971 In p r e s e n t i n g t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f the requirements f o r the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r re ference and s tudy . I f u r t h e r agree t h a t pe rmiss ion fo r e x t e n s i v e copying o f t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department o r by h i s r e p r e s e n t a t i v e s . It i s understood that copying or p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l ga in s h a l l not be a l lowed wi thout my w r i t t e n p e r m i s s i o n . an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that Department of The U n i v e r s i t y o f B r i t i s h Columbia Vancouver 8 , Canada ABSTRACT A study involving both the applications of internuclear double resonance (INDOR) spectroscopy and the synthesis of carbohydrates con-taining organometallic substrates was undertaken. The thesis is divided into two sections to f a c i l i t a t e discussion. The f i r s t section involves an evaluation of homonuclear and heter-onuclear INDOR as a tool for spectral analysis. Homonuclear INDOR studies of carbohydrate derivatives such as 2-deoxy-a-D-arabino-hexo-pyranose(l), sucrose octaacetate(2) and D-ribose(3) have shown this technique to be extremely effective for analysing multi-line spectra having hidden transitions. Within certain power-level ranges, the res-olution of the INDOR. responses is comparable to that of the normal pro-ton spectra. Heteronuclear INDOR spectra were obtained indirectly by monitoring specific transitions in the proton spectra while irradiating the appro-priate heteronuclear frequency. In this manner, the methoxy-methyl 13 C INDOR shift of a number of anomeric gluco- and xylo-pyranoside der-13 ivatives were measured. The 'C shifts of the 0£anomers were found to be ca. 2 ppm to high f i e l d of the p anomer. This difference is more diagnostic than the small 0.1 ppm variation between two methoxy-methyl anomers in the H n.m.r. spectra. Such diagnostic differences were not 13 found for the acetoxy-methyl C shifts of anomeric sugar acetate derivatives. Other heteronuclear INDOR spectra were readily obtained for organo-metallic compounds such as trimethylphosphite(15), trimethyltin chloride (16), bromosilane(17) and tetramethyl lead(18). The second section of the thesis deals with the synthesis of some phosphorus and organometallic derivatives of monosaccharides. Treat-ment of various primary O-tosyl sugar derivatives with lithium diphenyl-phosphine reagent ,19_, gave, in ca. 70% yields, such products as diphenyl 1^,2 :3,5-di-0-methylene-ct-D-glucofuranosej 6-C-phosphine oxide(20b), diphenyl ^ methyl 2,3,4-tri-fJ-acetyl-a-D-glucopyranosidej 6-C-phosphine oxide(21b) and diphenyl i*l,2:3,4-di-0-isopropylidene-a-D-galactopyranose: 6-C-phosphine oxide(22b). The facile preparation of 22b is rather unique because of the known resistance of the O-tosyl group of the sugar reactant toward nucleophilic displacement. Diphenyl tp,6-anhydro-l,2-O-isopropylidene-a-L-idofuranosej 5-C-phosphine oxide(24b) was obtained in 31% yield via nucleophilic displacement of the corresponding 5-0-mesyl derivative using 19_. This typified the much lower yields obtained from secondary sulphonyl displacements. Other secondary sugar diphenyl-phosphine oxide derivatives were prepared in 50-70% yields from the opening of sugar epoxides with 19_. Notably, the scission of the epoxide bond of methyl 2,3-anhydro-3-L-ribopyranoside(29) gave diphenyl ^methyl 2,4-0-acetyl-B-L-xylopyranosidej. 3-C-phosphine oxide(30). The product was found to favour the conformation in which a l l the major substituents are axial. 31 The P chemical shifts for a l l the sugar diphenylphosphine oxide , derivatives were measured using the INDOR technique. These shifts pro-vided confirmation that the products were phosphine oxides. It was fur-31 3 ther shown by use of P decoupling that Jp ^ ranges of 6.6-11.8 Hz and 22-33 Hz for 60° and 180° dihedral angles respectively, were present in these products. These couplings are similar to those reported for, vicinal P-H couplings in hydroxy phosphonate derivatives. Preliminary experiments, undertaken to evaluate the effectiveness of other organometallics such as lithium triphenyltin(33), lithium t r i -phenyllead(34) and lithium triphenylsilane (35) , showed 33_ to be an effec-tive reagent for displacing primary O-tosylates and opening epoxides of sugar derivatives. Both 34 and 35 gave poor yields when used to dis-place the 0-tosylate group of 20b. Generally, organometallic species were shown to provide a viable means of synthesizing novel sugar : derivatives. TABLE OF CONTENTS SECTION 1 INTRODUCTION 2 RESULTS AND DISCUSSION . 12 A. 1H-[1H] INDOR 12 B. '^H-t^C] INDOR 24 C. Some 1H-[M] INDOR spectra 34 EXPERIMENTAL .' ,44 A. 1H-[1H] INDOR experiments .44 B. ^ - [ ^ C ] INDOR experiments... .....46 C. Other heteronuclear INDOR experiments .53 SECTION 2 INTRODUCTION. . ,56 RESULTS AND DISCUSSION 62 A. Preparation of diphenylphosphine reagent .62 B. Reaction of primary O-tosylates 69 C. Reactions of secondary tosylates and epoxides...... ..85 D. Stereospecific dependence of Jp ^ . .j. .100 E. Reactions of some LiM0^ reagents with sugars ..106 EXPERIMENTAL. .119 APPENDIX A. : .' .138 APPENDIX B 145 REFERENCES 147 LIST OF TABLES Table 1 H and C N.m.r. parameters for methoxy-methyl • resonances of pyx-ariose derivatives 31 13 Table 2 C N.m.r. parameters for acetoxy-methyl reson-ances of pyranose derivatives 32 1 13 Table 3 Resonance frequencies for the H and C reson-ances of tetramethylsilane 51 Table 4 Results of adding lithium diphenylphosphine(19) to 1,2: 3,5-di-0_-methylene-6-0-tosyl-a-D-gluco-furanose (20a) . 7. . . 68 Table 5 Product yields obtained from the treatment of primary sugar tosylates with 19_ 72 Table 6 Chemical shifts for primary diphenylphosphine oxide monosaccharides ' r 78 Table 7 Coupling constants for primary diphenylphosphine oxide monosaccharides ....... 1 79 Table 8 N.m.r. data for ethyl diphenyl-phosphorus deriv-atives in various valence states 81 Table 9 Chemical shifts of secondary diphenylphosphine oxide monosaccharides 89 Table 10 Coupling constants for secondary diphenylphosphine oxide monosaccharide 90 Table 11 Chemical shifts for various sugar derivatives containing metallic substituents 112 Table 12 Coupling constants for various sugar derivatives containing metallic substituents ...113 (v i i i ) LIST OF FIGURES Figure 1 1 H and INDOR spectra of 5-hydroxy-l,2,3,7,7 hexachloro~bicyclo[2 . 2 . l]hept-2-ene 9 Figure 2 Partial Hi •n.m.r. and INDOR spectra for a f u l l y mutarotated solution of 2-deoxy-a-D-arabino-hexopyranose (1) T 14 Figure 3 . (A) Partial "*'H n.m.r. spectrum of sucrose octa-acet.ate (2) . (B,C,D, and E) INDOR responses obtained from monitoring transitions 1,8 3 and 5 fespec- • tively 17 Figure 4 *H n.m.r. spectrum and INDOR responses obtained for a f u l l y mutarotated solution of D-ribose 20 Figure 5 *H n.m.r. spectrum of methyl 2,3,4-tri-O-acetyl-B-D-xylopyranoside (7) 27 13 Figure 6 C .INDOR. spectra of the methoxyl-methyl carbon of (A) a^S^^tri-Oj-acetyl-e-D-xylopyraiiosideCT) and (B) 2,3,4-tri-O-acetyl-a-B-xylopyranoside 28 .Figure 7 *H n.m.r. spectrum of trimethyl phosphite (15_) .......... 36 Figure 8 3 1 P INDOR spectrum of trimethyl phosphite(15) .37 Figure 9 *H n.m.r.' spectrum of trimethyltin chloride(16) 38 i]7 119 Figure 10 (A) Sn and (B) Sn INDOR spectra of t r i -methyltin chloride (16) 39 Figure 11 " Si INDOR spectrum of bromosilane(17) ..41 207 Figure 12 Pb INDOR spectrum of tetramethyl lead(18) .42 Figure 13 (A) Partial H. n.nijj. spectrum of 22b; (B) spectrum with "P decoupling 75 Figure 14 3 1P INDOR responses for 22b_ ' 76 Figure 15 (A) Partial H nan^r. spectrum of 24b; (B) spectrum with P decoupling 87 Figure 16 Partial *H n.m.r. spectra of organometallic derivatives of 1,2 :3,5-di-0-isopropylidene-ct-' D.-glucofuranos'e I l l Figure 17 INDOR responses for H„ resonance at (A) 100 (B) 250 and (C) 500 seconds 144 LIST OF FLOW SHEETS Flow Sheet 1 30 Flow Sheet 2. 71 Flow Sheet 3 88 Flow Sheet 4... 109 ACKNOWLEDGEMENT Much gratitude is expressed for the encouragement and advice offered by Dr. L. D. Hall during the course of this work. Special thanks go to David Jones for i n i t i a t i n g my interest in INDOR spectroscopy. To my wife, Luisa, whose patience and understanding were pressed to their limits during the typing of this work, I respectfully dedi-cate this thesis. PREFACE The development of analytical methods for structural deter-minations has had a significant impact on the course of research in both organic and inorganic chemistry. New instrumentation has often ' provided the impetus for the synthesis of larger and/or more "exotic" molecules. The carbohydrate f i e l d , like many other chemical disciplines, has been enormously influenced by the nuclear magnetic resonance (n.m.r.) technique. The last decade has seen an interesting relationship evolve be-tween the two disciplines. On one hand, carbohydrate derivatives are often used as model systems for studying the stereospecific depend-encies of n.m.r. coupling constants, while on the other, n.m.r. spec-troscopy serves as the most widespread tool for elucidating the struc-tures of unknown monosaccharide derivatives. This thesis shows that a combination of conventional carbohy-drate chemistry and organometallic techniques, together with extensive applications of n.m.r. methods have made possible the synthesis and struc-tural assignment of several series of rather novel carbohydrate deriva-tives. These derivatives s t i l l retain the shape of a f a i r l y well-defined monosaccharide system arid thus provide n.m.r. infomiation which may be applicable to other chemical areas. In addition, this study further supports the philosophy that vari-ous branches of chemistry do not have to be treated separately but can be overlapped to increase general knowledge. For convenience, the thesis is divided into two main sections. The f i r s t section is primarily concerned with n.m.r. spectroscopy and deals specifically with the potential of intemuclear double resonance (INDOR) as a method for studying the structures of complex molecules. Partic-ular emphasis is placed on studies of carbohydrates and organometallic substances. The second section contains a discussion of the synthesis, via organometallic reagents, of several new classes of carbohydrate derivatives. It also illustrates the way in which n.m.r. spectroscopy, including the INDOR technique, can be used for structural assignments of these monosaccharide derivatives. In this sense, the latter section summarizes the main areas under consideration. SECTION I AN EVALUATION OF HOMONUCLEAR AND HETERONUCLEAR INDOR N.M.R. SPECTROSCOPY. INTRODUCTION The increasing application of nuclear magnetic resonance (n.m.r.) spectroscopy for determining configurations and conformations of organic compounds has made workers more aware of some inherent d i f f i c u l t i e s in this technique. For example, even with simple molecules containing only three mutually coupled protons, such as vinyl chloride"'", the n.m.r. * spectra can be of such complexity that simple f i r s t order analysis is impossible at 60 MHz.. Such spectra are commonly referred to as second order systems and the desired data are usually extracted by approxima-tions or by an extensive computer analysis using iterative techniques. Workers who apply n.m.r. spectroscopy as a tool for structural elu-cidation are often not interested in detailed analysis of spectra but instead only wish to extract the desired information in the easiest and most practical manner. Since most organic and inorganic chemists f a l l into this category, much stress has been placed on the development of ** newer n.m.r. methods which simplify spectral analysis. Many major n.m.r. advancements have taken place in the area of ; 2 double resonance . Through the use of field-sweep or frequency-sweep *The term f i r s t order analysis indicates that the chemical shifts and coupling constants are taken directly from the spectral splittings. **Interestingly, almost a l l these newer techniques were developed by physical chemists using rather small molecules and have only been applied to other more complex molecular systems after a long residence in the chemical literature. * systems , a secondary radiofrequency f i e l d of sufficient power can be used to remove a specific spin-spin coupling from an n.m.r. spectrum. This technique, referred to as spin decoupling, has certain restrictions in the case of homonuclear proton experiments. For instance, the i r -radiated proton must have a coupling constant which is not too large (<15 Hz) and the chemical shift of this proton must not be near that of the observed proton. Further, i f a proton spectrum is complex due to the number of nuclei present, the information normally obtained from a spin decoupling experiment may be hidden under overlapping resonances. For heteronuclear decoupling experiments much larger couplings can be irradiated. Judicious use of spin decoupling can provide one with exten-sive information on the "relative interactions" between different nuclei in a molecular system. Even though substantial applications and much theory have been pub-3 4 5 lished about spin, decoupling ' ' , few researchers truly understand the process occurring here. For cases where the normal spin decoupling experiment is restricted, the decoupling power can be decreased to a point where only a single transition of a resonance is irradiated. This process, commonly referred to as spin t i c k l i n g , perturbs the transitions connected to the energy levels in common with the irradiated transition. Experimentally, * A field-sweep experiment uses a constant radiofrequency for ob-servation while the resonance of interest is brought into resonance by scanning the magnetic f i e l d . In a frequency-sweep experiment, the f i e l d is held constant while an observing radiofrequency is swept through a portion of the spectrum. It is most convenient to use the latter method for spin decoupling or spin tickling experiments. one observes this as a sp l i t t i n g of some lines in the spectrum, the size of the s p l i t t i n g being dependent upon the power of the irradiating f i e l d . Spin t i c k l i n g , however, can also suffer from problems of reson-ance overlap. In many instances i t would be desirable to remove these overlapping resonances and study only a particular portion of the spectrum at a time. This section of the thesis w i l l examine the application of InterNuclear DOuble Resonance (INDOR) to achieve such subspectral an-alysis. It w i l l be shown that INDOR can be readily applied to simplify 31 complex proton spectra and also can be used to observe indirectly P, 13 19 117 119 207 C, S i , Sn, Sn, and Pb spectra of molecules containing these heteronuclei . INDOR*, as the f u l l name implies, is essentially a double reson-ance technique. However, instead of sweeping a relatively weak observing radiofrequency, w^,(as in a frequency-sweep spin t i c k l i n g experiment) through a region of interest while irradiating a portion of the spectrum with another radiofrequency, > the function of these two radio-frequencies is essentially reversed. In the INDOR mode of operation, is used to monitor a particular transition in the spectrum while w is scanned through the spectrum. This situation is relatively easy to ac-complish on a Varian HA-100 instrument (see Experimental Section). When CL>2 corresponds to a transition which is connected to the monitored line, the intensity of this monitored line w i l l change. Such changes can be *This name was chosen by Baker as a partial analogy to the ENDOR^  experiment in electron spin resonance spectroscopy. conveniently recorded on the spectrometer, pen carriage. Discussion of the types of responses produced w i l l be considered further on. The f i r s t INDOR experiment was reported by Baker in 1962. By moni-13 toring a C s a t e l l i t e line in the fluorine spectrum of trifluoroacetic acid at 56.4 MHz while sweeping another radiofrequency f i e l d through the range 15.090-15.092 MHz, he was able to indirectly reproduce the quartet 13 comprising the C spectrum of the trifluoro-methyl group. In a later 7 19 19 paper , Baker demonstrated how F—[ F] INDOR could be used to assign lines and to determine the relative signs of the coupling constants for a mixture of erythro and threo isomers of trans l,4-dichloro-3,4-dibromoperfluorobut-l-ene. Several other, groups have become active in applying the INDOR method. The basic research has been divided into two areas. The f i r s t area 8 1 1 involves homonuclear techniques , usually H—[ H]. In this context, 8c Kowalewski has made extensive use of INDOR to determine the relative * signs of highly coupled proton systems . A namesake used this method to measure small ( 0.1 Hz ) n.m.r. s p l i t t i n g s ^ . Although much success had been achieved in this area, only relatively simple molecules had previously been studied, and i t appeared that the application of this technique to more complicated structures was being overlooked. This ap-parent oversight encouraged our i n i t i a l ' interest in INDOR and ultimately led to successful applications in the analysis of carbohydrate spectra. A system of nuclei whose coupling constants and chemical shifts are such that they produce a second order spectrum is referred to as a highly coupled system. Recently, more interest has been devoted to the second area, that of heteronuclear INDOR. These experiments require more sophisticated instrumentation than the homonuclear cases. One of the more popular 9 13 13 heteronuclei to be examined by this method is C. The . C spectrum ... 13 can be observed indirectly by monitoring the C s a t e l l i t e lines of a *H n.m.r. spectrum while scanning an irradiating radiofrequency f i e l d 13 13 through the C resonance region. In this manner, the C spectrum is recorded at the sensitivity of protons which is ca. sixty-four times 13 that of the directly observed C spectrum. One drawback in this method 13 1 of observation is that the C- H satellites to be monitored in the proton spectrum must be clearly visible in order to perform the experiment. n + u i • i 1 9 9 u H 9 c 1 9 5 n < - 2 9 c - 3l„ , 103„, Other nuclei such as . Hg, Sn, Pt, S i , P, and Rh have also been studied by heteronuclear INDOR.^ ^ The extensive application of double resonance in recent years has resulted in a deluge of literature directed at the theory behind this phenomenon^'^ Since the INDOR mode is just another way of recording a double resonance experiment, most of this theory can be applied to INDOR observations. Several models have been presented to qualitatively explain some of the various possible observations (see Appendix A). Needless to say, although these models are somewhat simplified the help they provide chemists, who will be primarily concerned with synthesis; and structural analysis, make their presentation useful. Since this , study is primarily concerned with the applications of INDOR to specific structural problems in chemistry, detailed theoretical discussions w i l l not be considered here. An excellent review article by Hoffman and Forsen presents a com-prehensive discussion of multiple resonance from both a theoretical and pragmatic point of view. More recently a discussion of heteronuclear 1 (5 17 magnetic resonance has been presented by McFarlane , while Kowalewski has written a small article reviewing INDOR applications. * An INDOR experiment can produce two main categories of responses, (see Appendix A). If the irradiation frequency, tn^, is set at a power level so that both population and energy level changes occur in the system (changes which are similar to those occurring in a spin-tickling experiment) a l l INDOR responses are negative-going. On the other hand, INDOR responses can be either positive- or negative-going i f the power of is set at a lower level so that population changes dominate. * i As of late, there has been much confusion over the terms ^ ^ "overhauser experiment" and "transitory selective irradiation (TSI) ' experiments" leading to an interchange of these expressions by some researchers. In an overhauser experiment, a transition is irradiated continuously causing a redistribution of population which depends on both energy level arrangements and relaxation rates in the spin system. This effect can occur and cause interpretation problems in an INDOR experiment when the sweep rate is sufficiently slow so that saturation arises re-sulting in secondary distribution of populations throughout the energy levels. Only i f the spin lattice relaxation time of the irradiated transition is much shorter than that of the monitored transition, do these secondary changes become minimized in the overhauser experiment. In the TSI experiment, the population changes arise from transitory irradiation of selected lines. Here, intensity changes are independent of the relaxation process, provided that the experiment is short in comparison with thermal relaxation times. Most frequency-sweep INDOR experiments occur under the conditions of adiabatic fast passage and thus f a l l into the TSI category. I n g e n e r a l , the former case predominates f o r heteronuclea'r INDOR • I : experiments, w h i l e the l a t t e r s i t u a t i o n u s u a l l y occurs .in .11—[•!!] INDOR ob s e r v a t i o n s . In some in s t a n c e s both of these s i t u a t i o n s can occur simul t a n e o u s l y . In the case where p o p u l a t i o n changes predominate; a p o s i t i v e respon r e s u l t s when the monitored a n d . i r r a d i a t e d t r a n s i t i o n s are connected such t h a t they e f f e c t i v e l y span two quantum jumps ( p r o g r e s s i v e t r a n s i t i o n s ) as shown i n |A J. I f these two t r a n s i t i o n s are connected so that they terminate on l e v e l s w i t h the same magnetic quantum number ( r e g r e s s i v e t r a n s i t i o n ) , as shown i n y>g, a negative response occurs. Consider an INDOR experiment performed on 5 -hydro 'xy - l ,2 ,3 ,7 ,7 , hexa • c h l o r o - b i c y c l c [ 2 . 2 . 1 ] h e p t - 2 - e n e * ( F i g . l ) . The normal 100 MHz proton •A-Sample provided by Mr. R. B. Malcolm, Chemistry Department, U.B.C. FIG. 1 H and INDOR s p e c t r a - o f 5 - h y d r o x v - l , 2 , 3 , 4 , 7 , 7 , h e x a c h l o r o -b i c y c l o [ 2 . 2 . 1 ] h e p t - 2 - e n e i n CDC 1 Q s o l u t i o n . spectrum (middle of Fig. 1) is essentially that of a f i r s t order AMX system. The hydroxyl hydrogen atT6.7 is undergoing fast exchange and does not exhibit couplings with the other nuclei in the molecule. The responses obtained from the monitoring of lines 12 and 5 while sweeping the rest of the spectrum with an irradiating f i e l d are shown respectively in the upper and lower portions of Figure 1. When the monitoring and irradiating frequencies coincide, a characteristic beat pattern results. The sharp responses can be assigned unambiguously to a particular tran-sition in the normal spectrum. It should be noted that some minor broad responses are also present in the INDOR spectra of Figure 1. These un-desirable effects can arise in two ways. The monitoring frequency may, not reside exactly on the transition which is to be monitored, resulting in a positive or negative spike appearing on either side of the INDOR 19 response. Otherwise the power of the monitoring f i e l d may be of a higher level than is desirable, causing secondary distributions of pop-* ulation. For these particular spectra the former situation produced -, the unwanted side effects. Needless to say, they can easily be differen-tiated from the true INDOR responses. The INDOR spectra of this * Ideally the monitoring frequency should be of such a bandwidth that only one transition is irradiated at a time. This is especially d i f f i c u l t to achieve with closely spaced transitions £2 Hz apart). Often a com-promise must be made between the intensity of the INDOR responses and the degree of secondary responses. bicycloheptene derivative also show the relative signs of the couplings * t 0 b e JAM^ JAX = JMX a S e x P e c t e d -Having demonstrated the relative ease of the INDOR technique on ' this somewhat t r i v i a l system, the question now arises whether this method can be successfully applied to more complex systems. As the remaining portion of this section w i l l demonstrate, the INDOR technique shows promise as a tool in the analysis of proton spectra of larger molecules. Indeed, in the case of heteronuclei i t provides an inexpen-sive way of obtaining a multi-nuclei spectrometer. See Appendix B for sign determinations and energy level diagram. RESULTS AND DISCUSSION A. 1H-1H INDOR As previously mentioned, most organic and inorganic chemists use proton n.m.r. to help to elucidate the structure of their compounds. Carbohydrates, because of the various electronegative substituents which can be substituted, exhibit a variety of proton shifts resulting in spectra which range from simple f i r s t order cases to highly second order situations. Even in the most complex spectra, the anomeric proton (C—1) usually resides to lowest f i e l d (excluding benzylic and aromatic protons). Thus, the anomeric proton resonance can be readily monitored to produce INDOR responses for mutually coupled nuclei. Since the INDOR experiment requires a few well defined transitions to monitor i n i t i a l l y , the carbohydrate f i e l d was chosen as the area to f i r s t investigate. HO OH H 1e The proton spectrum of 2-deoxy-D-arabino-hexopyranose[I] in solution indicates a mixture of a and $ anomers ( see diagram |JQ|). 20 Previous studies by Lenz and Heeschen reported chemical shifts of the anomeric protons and coupling constants for H.^  and 11^. Hall and 21 Manville investigated the spectral region of this sugar using spin decoupling; they were able to completely assign the subspectra of this region for each anomer. The remaining protons are lumped together in an unresolved multiplet between T5.5 and 6.5. The normal Hj and resonances of a f u l l y mutarotated solution of 2-deoxy-arabino-hexopyranose in solution (ca. 30% w/ y) a r e shown in the middle section of Fig. 2. The two H^  resonances can be assigned to the appropriate anomer by applying to J H „ the well known Karplus 1 ' 2 22 relationship relating coupling constants to dihedral angles. Using the INDOR technique, one can easily assign the multi-line H^  protons to the appropriate anomers. Figures 2A, 2B, 2C and 2D show the INDOR responses obtained by monitoring transitions 1,2,3,4 respectively (the 6[H, ] anomeric proton) while scanning the rest of the spectrum. As can be seen, the INDOR spectra are well resolved and both positive and negative responses occur indicating that population changes are the dominant mechanism for these INDOR spectra. Two pairs of spectra (Figs. 2A and 2D or 2B and 2C) each serve to identify the positions of a l l sixteen of the BfH^] transitions and give the assignment shown in the upper portion of the normal spectrum (Fig. 2) . Similarly the a[H~] 700 750 8.00 7.00 750 ; 8.00 FIG. 2 . P a r t i a l H n.m.r. and INDOR spectra f o r a f u l l y mutarotated so l u t i o n of 2-deoxy-a-D-arabino-hexopyranose(l) i n D-0 (see t e x t ) . transitions can be assigned from monitoring lines 8,7,6 and 5 to give the spectra shown in Figs. 2E, 2F; 2G and 2H respectively. Again, Figs. 2E and 2H or 2F and 2G provide sufficient information to complete the assignment of the ctf^] resonance. Noticeably, the quality of the INDOR spectra obtained from moni-toring the four a[H^ g] lines of the a anomer is much poorer than that of the others in Fig. 1. There are two related reasons for this. F i r s t , ^2e 1 * n t n ^ s m ° l e c u l e 1 S small (0.7 Hz), requiring the observing and decoupling fields to be significantly decreased in order to prevent i r -radiation of more than one transition at a time. In these situations there must be a compromise between seriously decreasing the signal to noise ratio by decreasing the strength of the irradiating radiofrequency f i e l d and the partial irradiation of other transitions by using too strong a f i e l d . The latter creates "ghosting" in the INDOR spectra. Usually these "ghost lines" can be tolerated because of. their low inten-s i t y compared to the normal INDOR responses. Secondly, to obtain the sharpest INDOR responses the observing radiofrequency (w^ ) should be set to within 0.1 Hz of the monitored line. In this case some of the °( Hj e transitions are not sufficiently well separated to easily accomplish this limit. This results in a broadening of the observed responses together with the formation of "spikes" on either side of the major re-sponse. As with the spin ti c k l i n g experiment, the INDOR responses provide sufficient information to determine the relative signs of the coupling constants; specifically, assuming J.._ is positive then J_ and i-iC-H Z e > 1 J„ are positive while J~ „ is negative as expected. With the establishment of the ease of operation and the sharpness of the INDOR responses in the case of 2-deoxy-D-glucose, i t was of inter-est to see i f this technique had potential in more complex sugar systems. The proton n.m.r. spectra of oligo- and poly-saccharide derivatives are complex and spectral assignments are generally d i f f i c u l t . Recent 23 studies with superconducting solenoids (at 220 MHz) have provided some improvement over earlier 100 MHz studies but are s t i l l not com-pletely satisfactory. The proton spectrum of sucrose octa-acetate(2) at 100 MHz is shown in Figure 3A. Several of the transitions can be readily assigned; however, complete analysis of either the furanose or pyranose ring.pro-tons is d i f f i c u l t by normal inspection. Starting from the doublet at x4.32, which can be easily assigned to the anomeric proton of the glucopyranose ring, the position of a l l . the pyranose ring protons can be found by a series of INDOR experiments. Monitoring transition 1 produces INDOR responses which assign the resonance (Fig. 3B) . Then, by monitoring line 8 of the quartet, both H^  and H^  transitions give INDOR responses (Fig.- 3C) since they satisfy the connectivity relationship in the energy level system. Monitoring lines 5 (Fig. 3D) and 3 (Fig. 3E) produces INDOR responses for identical transitions in the spectrum; the only difference occurs in the direction I. . . . I • I • I ••• I • 1 • 1 . • 1 1 4.0 4.5 5.0 5.5 6.0 T FIG. 3. (A.) P a r t i a l H n.m.r, spectrum of sucrose octa-acetate(2) i n acetone-d with assignment of- pyranose protons obtained from INDOR spectra. (B,C,D, and E) Responses obtained from monitoring t r a n s i t i o n s 1,8,3 and 5 r e s p e c t i v e l y (see t e x t ) . of these responses. This situation arises because of the nature of the two lines monitored; one is an a transition, while the other is a * 3 transition of the same proton. The low f i e l d INDOR responses of these latter two spectra correspond to transitions, thus inden- . ti f y i n g the t r i p l e t comprising lines 3,4 and 5 as the resonance and also assigning the upfield response as belonging to the H protons of 5 the glucopyranose ring. The complex responses for H<- arise from the Effectively, the designation of a transition as a or 3 depends upon the relative order of the energy levels connected by this transition. For instance, consider an energy level diagram for an AX system. Of a P a P a P P Transitions 1 and 2 both involve a change in the spin of the X nucleus from the 3 to a state and form the two lines of the X doublet in the AX spectrum. However, these two transitions d i f f e r . In the case of transition 1, the A nucleus is always in the 3 state, while for tran-sition 2, the A nucleus is always in the a state. A r b i t r a r i l y , one can label transition l'the 3 transition and transition 2 the a tran-sit i o n . If the two middle energy levels now had their order reversed, .the positions of the a and 3 transitions in the AX spectrum would be interchanged. highly coupled relationship between the and protons and are not assignable on a f i r s t order basis. The results of these sequential INDOR experiments is to produce the assignment shown at the top of Fig. 3A. It is noteworthy that the resonance is essentially hidden in the nor-mal spectrum but can be assigned easily using INDOR. No attempt was made to determine the furanose positions since the i n i t i a l transitions, to be used as a starting point for spectral analysis, could not unequi-vocally be assigned by inspection. This further stresses an important prerequisite for performing these experiments; the transitions from at least one assignable proton should be well separated from the bulk of the spectrum in order to be used as a point of embarkation for moni-toring INDOR responses. 24 Lemieux and Stevens " have previously demonstrated the proton n.m.r. spectrum of the free sugar, D-ribose(3), in aqueous solution to be a band of complex multipets as shown in the middle of Fig. 4. A f u l l y mutarotated solution of D-ribose consists of a mixture of four anomers (shown in . FIG. 4. H spectra f o r a f u l l mutarotated s o l u t i o n of D-ribose(3) i n using the methyl groups o f sodium dimethylsilapentane sulfonate f o r an i n t e r n a l lock s i g n a l . The normal spectrum i s shown across the top of the Figure with some of the i n t e n s i t y between x6.0 and 6.5 due to the methylene protons of the lock standard. The H-2 responses obtained from monitoring l i n e s 2,3,5, and 7 are shown i n A,B,C, and D r e s p e c t i v e l y . Using kinetic data and the magnitude of coupling constants, the four low f i e l d doublets can be assigned to the anomeric resonances of the a-24 25 and 8-furanoses and a- and 3-pyranoses. ' The INDOR spectra which are the result of monitoring lines. 2,3,5 and 7 are shown respectively in Figs. 4A, 4B, 4C and 4D. These INDOR responses serve to identify-the H resonances of each anomeric form and confirm the former assign-ment of these resonances obtained by field-sweep spin decoupling ex-24 periments. It should be added that some of the broad responses evi-dent in some of the spectra is due to the resolution of the spectro-* meter being somewhat unstable during the experiment. The advantage of seeing responses for only transitions connected to an energy level in common with the monitored transition is greatly evident in these experiments. That a proton exhibits a coupling with other protons is evidence of such a connectivity relationship. Furthermore, even though the molecules studied have previously been examined in the literature, the ease with which the INDOR mode of oper-ation provides this information demonstrates the high potential of this technique. Already INDOR applications have been extended to studies, of 2 6 configurations and conformations in the area of steroids and antibiotics * Using the TMS capillary inside the n.m.r. tube requires the sample to be spun faster than normal, probably resulting in much of the in s t a b i l i t y present in the spectrometer lock channel. Routinely, our laboratory now uses INDOR for most spin tickling or partial decoupling, experiments. At this point i t would be beneficial to consider some of the limit-ations which may be. encountered during INDOR studies of complex molecules. (a) Monitoring transitions in a highly coupled system:- This may produce a multitude of INDOR responses which cannot be assigned directly by inspection. Often the spectrum can only be solved using com-puter analysis of the spin system. (b) Monitoring a degenerate transition:- This w i l l produce more responses than i f a single, non-degenerate transition is being moni-tored and i t may be d i f f i c u l t to determine the multiplicity of the. INDOR spectrum. (c) Irradiating a transition at a power-level such that both  the response "mechanisms" can operate simultaneously:- It is possible under these conditions for a positive- and negative-going response to; effectively cancel each other out resulting in an INDOR spectrum which contains too few responses. (d) Monitoring a transition having an unusually long relax- ation time:- This condition may produce very broad responses. By in-creasing the scan rate this situation can be alleviated; however,,loss of resolution results. Fortunately, this does not appear to be a s e r i -ous problem in *H—[*H] experiments, although several examples have been encountered during heteronuclear INDOR measurements. (e) Monitoring a line in a system containing one or more  intense resonances:- A spurious response often is observed when the irradiating f i e l d is scanned through an intense response (i.e. methoxyl resonance) even though this particular resonance has no "connectivity" with the monitored line. Such a response arises as the result of over-loading certain HA—100 amplifier circuits. This condition can be readily avoided by operating at lower signal levels, either by sample dilution or by reduction of the "receiver gain" setting of the V-4311 unit. Often with higher molecular weight substances a compromise must be made with these spurious signals in order to detect the true INDOR responses. Alternately, computer time averaging can be used to intensify the responses. B. 1 H - [ 1 3 C ] INDOR An essential condition for achieving resonance in any n.m.r. experi-ment is satisfying the relationship a). = yH I o where and H q represent the radiofrequency and magnetic f i e l d respect-ively, while Cmagnetogyro ratio 3 _ magnetic moment Y " (for a particular isotope) ~ angular momentum 12 Fortuitously, the existence of C as a dominant isotope having zero magnetic moment has proven a boon to "''H magnetic resonance spectroscopy. This isotope exhibits no n.m.r. condition because the magnetogyro ratio 1 12 is zero; thus greatly simplifying the H spectrum since C - H couplings are nonexistent. 13 The C isotope,, present in 1.1% of naturally occuring carbon, has an associated nuclear magnetic moment with a nuclear spin quantum num-ber of 1/2. Due to the minor presence of this "n.m.r. active" species, i t provides l i t t l e interference with the normal proton spectra and has until recently been essentially ignored as a diagnostic species. With the development of more sensitive instrumentation, the importance of 13 using C n.m.r. spectroscopy to examine the molecular carbon skelton 28 . became evident. Most experiments have involved using either enriched samples or large diameter sample tubes together with computer time averaging tech-13 niques to observe C spectra. More recently the- interactions between protons and C have been removed by proton noise decoupling resulting in a corresponding increase in the line intensities which together with 13 29 an Overhauser effect further enhances the C signal. 13 The large chemical shift range of C (ca. 300 ppm) compared to 13 that of protons (ca. 10 ppm) along with the much larger C—H couplings 13 constants, clearly show the potential diagnostic advantages of Cl-over conventional *H- n.m.r. 1 13 H—[ C] heteronuclear INDOR can provide a less expensive and some-13 13 what more convenient way of observing C spectra provided the C 1 13 satellites in the H spectrum can be accurately monitored. The C spectrum is observed indirectly via the proton spectrum leading to en-6 9 hanced sensitivity ' (i.e. ca. 64-fold sensitivity increase) over con-ventional direct measurements, 13 C INDOR experiments are carried out in a similar manner to the *H—[*H] INDOR observations. The major difference arises from the need 13 to scan the C radiofrequency region (ca. 25.14 MHz using a 23,500 13 Gauss magnetic field) while monitoring a C s a t e l l i t e in the proton spectrum. A frequency synthesizer coupled to a digital sweep unit were 13 the means used in this laboratory to provide the irradiation of the C region. In addition, a frequency-frequency lock system was arranged be-tween the frequency synthesizer and the HA—100 radiofrequency unit, in order to provide the necessary s t a b i l i t y for accurate chemical shift determinations. The C INDOR technique has previously been applied to several 9 29 30 organic systems ' ' although not in the carbohydrate f i e l d . Direct 13 measurements of C spectra of carbohydrates have been carried out by 31 several groups. Some of these studies indicated that the chemical 1^  shifts of the anomeric methoxy-methyl 'C resonances have a stereo-specific dependence.^^>^ Since the satellites of the methoxy-methyl group in the proton spectra could be easily monitored, we under-13 took to further investigate this stereospecific dependence by C INDOR. These studies were then extended to O-acetyl-methyl and N-acetyl-methyl shifts. Figure 5 shows the *H n.m.r. spectrum of methyl 2,3,4-tri-O-acetyl-B 13 D-xylopyranoside(7) .. The high f i e l d C s a t e l l i t e of the methoxy-methyl and acetoxy-methyl resonances, recorded at much higher gain, are 13 shown as an insert. Figure 6A shows the C INDOR spectrum obtained by monitoring the methoxy-methyl group of 7_ (see Experimental Section). It is important to note that this spectrum represents a single scan 13 recording at moderate concentrations with natural abundance C. For 13 comparison purposes, the C spectrum of the methoxy-methyl carbon of, methyl 2,3,4-tri-0-acetyl-ct-D-xylopyranoside(6) is shown in Fig. 6B. As can be seen, this shift is significantly different from that of 13 7 (Fig. 6A). It should be noted that a l l of the C INDOR responses are negative-going, indicating that the "spin t i c k l i n g " mechanism ( see Appendix A),involving both population and energy level changes, is the main process operating here. This might well be expected from the highe FIG. 5 H Nuclear magnetic resonance spectrum (100 MHz) o f methyl 2 , 3 , 4 - t r i - 0 - a c e t y l - g - D - x y l o p y r a n o s i d e ( 7 ) i n CDC1 s o l u t i o n . The C s a t e l l i t e s of the metfioxyl-methyl and acetate-methyl s u b s t i t u e n t s , measured at h i g h e r g a i n , are shown as an i n s e r t . FIG. 6 Natural abundance C resonances measured by the H-[ C] INDOR technique; (A) the methoxyl-methyl carbon of methyl 2 , 3 , 4 - t r i 0-acetyl-B-D-xylopyranoside(7); (B) the methoxyl-methyl carbon of methyl 2,3,4-tri-0-acetyl-a-D-xylopyranoside(6). power l e v e l of the i r r a d i a t i n g f i e l d (compared to the *H—[*H] INDOR case) being used f o r these heteronuclear INDOR experiments. When attempts were made to lower the i r r a d i a t i n g power l e v e l i n order to predominantly produce population changes, the responses became broad and weak. 1 13 The H and C parameters of the anomeric methoxy-methyl resonances of compounds 4_-l_l_ are l i s t e d i n Table 1. The 1H s h i f t s are presented here to indicate the d i f f i c u l t y of using these values to d i s t i n g u i s h 13 between a and B anomers. The C values show a much larger s h i f t d i f f e r -ence, enough to be diagnostic i n determining the anomeric configuration 13 of the methyl glycoside. In a l l cases, the C s h i f t value of the'B anomer i s to higher frequency (lower f i e l d ) than that of the a anomer. Further, the s h i f t differences seem to be s l i g h t l y larger f o r the methyl glycosides than f o r the methyl glycoside peracetates. 13 1 The C-H coupling constants f o r the methoxy-methyl substituents were measured and are also l i s t e d i n Table 1. In view of the rather s l i g h t v a r i a t i o n s i n these values, they do not appear to have any diag-n o s t i c s i g n i f i c a n c e . 13 The C s h i f t s of the acetoxy-methyl resonances of several ace-t y l a t e d pyranose derivatives(5,12-14) are summarized i n Table 2. The 13 • C s h i f t s f o r 5_ and 1_2 a l l f a l l within a narrow range. Thus, i n con-13 t r a s t to the C•methoxy-methyl resonances, the acetoxy-methyls appear to have no diagnostic value ( i . e . no differences between a x i a l and equa-13 t o r i a l or primary and secondary p o s i t i o n s ) . Notably, the C s h i f t s of O A c O A c 2 4 R= H ; RzOMs 6 R=H; R=OMe 8 R= H ; R=OMe 5 R = O M e ; R=H 7 R=OMe;R=H , 9R=OMe;R=H 10 R=H.; R=OMe 12 R=OAc;R'=H 13 R=H; R = N H A C H R=OMe;R=H 14 R=NHAc ; R=H Flow Sheet 1 TABLE 1. H and C Nuclear magnetic resonance parameters f o r methoxy-methyl resonances of pyranose d e r i v a t i v e s . j *!! S h i f t s 1 3C S h i f t s J 1 3 1 C-H (Hz) Compound I i Isomer j xOMe ! AT I i i ! v 1 3C+ 6 1 3 G t j A 6 5 I ! 4 • 5-a " P " 6.590 | 1 0.080 .6.510 ; 25.147025 ! 55.12 25.147065' 56.70 1.58 148 143 6 7 a " P" 6.610 ': 1 0.075 6.535 ; 25.147029 25.147063 55.27 56.63 j 147 1.36 ; 145 8 a * P # 6.060 .5.870 0.190 25.147051 25.147108 56.43 ! • 144 i 2.26 j 58.69 j . ; 145 : i .10 11 a # P # 6.085 5.990 0.095 25.147056 25.147106 56.63 58.61 1.98 145 148 {xOMe - xOMeQ}. Chemical s l i i f t , i n Hz, measured at a magnetic f i e l d such t h a t the protons of t e t r a m e t h y l s i l a n e resonate at 100.000000 MHz. t Chemical s h i f t , i n ppm, r e l a t i v e to that o f t e t r a m e t h y l s i l a n e ; see Experimental Section and Table 3. It should be noted that a higher resonance frequency corresponds to a lower f i e l d chemical s h i f t . " In CDC1 3 s o l u t i o n . # In D 70 s o l u t i o n . Table 2: C Nuclear magnetic resonance parameters f o r the acetoxy-methyl resonances o f pyranose sugars. Compound Chemical S h i f t 1 3 r * v C 61 3 C f J 1 3 C - *H <Hz> 25.146157 20.59 130 25.146157 20.59 131 25.146160 20.72 130 25.146156 20.54 130 25.146159 • 20.67 131 12 4 25.146157 20.59 j 131 25.146157 (2x) 20.59 (2x) | 130 25.146156 20.54 1 131 13 § 25.146220 23.08 131 " t i ^ | 5 J 25.146217 22.93 131 * Chemical s h i f t , i n Hz, measured at a magnetic f i e l d such t h a t the protons o f t e t r a m e t h y l s i l a n r e s o n a t e at 100.000000 MHz. These data are g i v e n i n the order f o r which the ^ C - s a t e l l i t e s o f p r o t o n spectrum were monitored, which i s from low- to h i g h - f i e l d . T Chemical s h i f t , i n ppm, r e l a t i v e t o t h a t o f t e t r a m e t h y l s i l a n e , see Experimental s e c t i o n and Table 3. I t should be noted t h a t a h i g h e r resonance frequency corresponds to a lower f i e l d chemical s h i f t , t In CDC1, s o l u t i o n . § In D_0 s o l u t i o n . Table 2: C Nu c l e a r magnetic resonance parameters f o r the acetoxy-methyl resonances o f pyranose sugars. Chemical S h i f t - ' : Compound • 13- * v C 6 1 3 C f J 13 c- 1 H (Hz) 25.146157 20.59 130 5: 4 25.146157 20.59 131 25.146160 20.72 130 25.146156 20.54 130 i I 25.146159 20.67 ! 131 25.146157 20.59 131 12 + 25.146157 (2x) 20.59 (2x) 130 ; 25.146156 20.54 131 13 s 25.146220 23.08 131 ' i 14 s 25.146217 22.98 131 J * Chemical s h i f t , i n Hz, measured at a magnetic f i e l d such t h a t the protons o f t e t r a m e t h y l s i l a n e 13 resonate at 100.000000 MHz. These d a t a are g i v e n i n the order f o r which the C - s a t e l l i t e s o f the p r o t o n spectrum were monitored, which i s from low- t o h i g h - f i e l d . T Chemical s h i f t , i n ppm, r e l a t i v e t o t h a t o f t e t r a m e t h y l s i l a n e , see Experimental s e c t i o n and Table 3. I t should be noted t h a t a h i g h e r resonance frequency corresponds to a lower f i e l d chemical" s h i f t . * In CDC1 3 s o l u t i o n . . § In D o0 s o l u t i o n . the acetoxy-methyl resonances are cji. 1 KHz (40 ppm) to .lower frequency of the methoxy-methyl resonances. Besides this, the J..., are now ca. ^C-H 130 Hz, compared to ca_. 145 Hz for the methoxy-methyl resonances. In addition, the data for the acetate-methyls of the two acetamido derivatives(13,14) are reported in Table 2. Again, these shifts appear to have l i t t l e diagnostic value, although they are ca. 50 Hz (2 ppm) from the acetoxy-methyl shifts. 13 It i s evident from the above discussion that in some instances' C m.r. spectra can provide a good diagnostic tool for configurational 13 assignments. The advantages of measuring C m.r. spectra by the INDOR method rather than by direct measurements are: spectra can be obtained rapidly on a single scan basis due to.the increased sensitivity; spectra 13 of compounds containing natural abundance C at moderate concentrations 13 in 5mm O.D. n.m.r. tubes can be recorded; the C response obtained is directly associated with a "coupled" hydrogen in the *H spectrum thus serving to identify a specific carbon. Unfortunately, the major limita-tion requiring that the sat e l l i t e s in the *H n.m.r. spectrum can be moni tored, usually precludes using the INDOR technique to measure the ring carbons of sugars. The above discussion, has demonstrated the potential of INDOR to overcome the. low relative sensitivity of the *3C isotope. Similarly, INDOR spectra of any n.m.r. "active" species can be measured indirectly from the Hi n.m.r. spectrum provided that a coupling exists between'.a proton and the heteronuclus, M, in the molecule. Recently, INDOR spectra of several heteronuclei have been shown to be superior to those obtained 32 by direct observation. The greater natural abundance of most nuclear 13 spin 1/2 isotopes compared with that of C makes INDOR spectra of . 13 these nuclei generally easier to obtain than that of C. The following represent several examples of common isotopes which can be easily measured using the heteronuclear INDOR technique. . Most of the substances were chosen because they were easily obtained from suppliers. It should be noted that a l l compounds were measured using a single scan at natural abundance of solutions containing 30-40% w/v. Phosphorous is known to be an important element in chemical and biological systems. Even though i t exists in 100% natural abundance, 31 1 the low relative sensitivity of the P isotope (compared to the H 31 isotope) often makes direct P m.r. observations d i f f i c u l t . Recently, 31 33 INDOR has been used to obtain P m.r. spectra and to determine the magnitude and signs of *H—31P coupling constants . ^  The *H spectrum of trimethyl phosphite(15) , shown in Fig. 7, consists of a doublet with the 31 sp l i t t i n g equal to the P-O-C-H coupling. The P INDOR spectrum obtained from monitoring either of these two proton lines, while sweeping an 31 irradiating radiofrequency f i e l d through the P region, is shown in , 31 1 Fig. 8. Indirect observation of the P m.r. spectrum via a H resonance leads to a 16-fold sensitivity enhancement over direct measurement. 31 The P spectrum should consist of ten lines. However, the intensity ratio between the inner and outer pair of lines is 126 to 1. For this reason, the outer pair of lines, are not v i s i b l e . - It is noteworthy that the resolution of the lines is ca. 1 Hz at a sweep rate of 0.5 Hz/sec. arid the reproducibility of the chemical shift in Hz from tetramethyl-+ silane, resonating at exactly 100 MHz, is within - 1 Hz. In the organometallic and inorganic f i e l d s , t i n has always been an 35 31 117 important element. In contrast to P, the two isotopes, Sn and 119 Sn, are both species of low abundance with a nuclear spin of 1/2. Fig. 9 shows the Hi n.m.r. spectrum of trimethyltin chloride(16). The 119 doublets on either side of the main methyl peak are the Sn— H and 117 Sn-H coupled s a t e l l i t e transitions. Effectively, one is observing 117 three different molecular species in this spectrum: (CH,) Sn CI, 3 n o I -I q ^ (CH ) Sn CI and (CH ) Sn CI. Figs. 10A and 10B show respectively 3 3 117 119 the INDOR spectra obtained from monitoring one of the Sn and Sn satellites in the '''H n.m.r. spectrum. Again, the outer pair of lines are not visible because their relative intensity is very small compared with the middle lines. • - i — i — i — i • * * * * * * • • • • • • • • 1 • • 6.0 l-'iC. 7 !11 N.m.r. spectrum (100 MHz) of t r i m c t h y l phosphite(15) i n CDC1 3 s o l u t i o n . 10 H Z 40,486,434 Hz 31 FIG. 8 P INDOR spectrum of .trimethyl phosphite(15) i n CDC1 s o l u t i o n obtained from monitoring the u p f i e l d . l i n e i n the H n.m.r. spectrum. 35,637,547 Hz 6^  ,"Y\^^ 37,295,364 Hz 117, FIG. 10 (A) I W ? n ;ind (R) 1 1 9 S n INDOR spe c t r a o f t r i m e t h y l t i n c h l o r i d e (16) i n CDClj s o l u t i o n obtained from monitoring t h e i r r e s p e c t i v e s a t e l -l i t e l i n e s i n the H n.m.r. spectrum. The Si INDOR spectrum of bromosilane(17) is shown in Fig. 11. The slight dispersion mode characteristics of the signal are due to phase d i f f i c u l t i e s present during the measurement. Although the spectrum is relatively simple, the ease at which the quartet can be recorded pro-29 vides a convenient method of measuring Si chemical shift s . Recent 29 Si INDOR measurements on a series of s i l i c o n hydrides containing a variety of electronegative substituents have shown a large range of chemical s h i f t s . " ^ Finally, lead is another metal which has been used to some extent in 207 organometallic synthesis. Fig. 12 shows the Pb INDOR spectrum of tetramethyl lead(18) obtained by monitoring one of the s a t e l l i t e lines in the n.m.r. spectrum. In this case, the outer two pairs of lines are not visible because of their weak relative intensity compared with the inner lines. In summation, this investigation has shown that INDOR can be suc-cessfully applied to certain complex systems. The technique's advan-tage over conventional double resonance experiments arises from i t s selectivity. INDOR responses are obtained only for transitions satis-fying a connectivity relationship to the monitored transition in the energy level system. As such, the monitoring of a single transition produces INDOR responses for only a small number of lines in the n.m.r. spectrum. Essentially, the rest of the spectrum is not observed. Thus, 100 H Z 19,866,216 Hz 29 FICg 11. S i INDOR spectrum o f bromosilane(17) obtained by monitoring a " S i s a t e l l i t e l i n e i n the II n.m.r. spectrum. 100 H Z r 1 r 20,920,648 Hz 207, FIG. 12 Pb INDOR spectrum o f t e t r a m e t h y l lead(18) o b t a i n e d from m o n i t o r i n g the low f i e l d s a t e l l i t e l i n e H n.m.r. spectrum. each INDOR experiment breaks the normal spectrum down to a series of minor sub-spectra which can readily be analysed. Heteronuclear INDOR is just an extension of this technique to obtain, indirectly, spectra of nuclei other than *H at the relative sensitivity which is associated with *H. By a careful selection of the lines to be monitored in the *H n.m.r. spectrum, the entire heteronuclei spectrum can be constructed from the observed INDOR responses. EXPERIMENTAL The spectrometer system used for these experiments was a Varian HA—100 instrument with a high impedance 12" magnet. The INDOR instrum-ental modifications were carried out by Mr. Roland Burton of this de-partment. In the case of *H—[*H] INDOR, the modifications involved a 37 slight alteration in the unit described by A. W. Douglas. A. 1H-[1H] INDOR EXPERIMENTS 2-deoxy-D-arabino-hexopyranose(l) A detailed description, of the *H—[*H] INDOR experiments on 2-deoxy-D-arabino-hexopyranose (m.p. 156-158°) in D^ O solution w i l l be presented. Identical instrumental procedures were applied to the other compounds. A sample of 1_ (Pfanstiehl Laboratories, Waukegan, 111.) was lyo-philized several times with D2O. A solution of ca. 30% w/v was pre-pared and a capillary tube containing tetramethylsilane was used to pro-vide the lock signal. A normal frequency-sweep spectrum was recorded and the transition frequency of the anomeric resonances were measured on the Varian V-3415 counter to - 0.1 Hz. The HA-100 was then switched to the INDOR mode of operation which effectively substitutes an external oscillator for the normal observing sweep oscillator. The latter now becomes the oscillator providing the irradiating frequency,^. In these experiments a fre-quency synthesiser was used as the external oscillator because of the high s t a b i l i t y of i t s output frequency. Using an attenuation of 26 db on the proton radiofrequency unit and setting the spectrum amplitude at maximum, sweep oscillator set at 0.6-0.9 (as indicated by the dial) and external oscillator at 10-20 mv, transition 1 of Fig. 2 was moni-tored. I n i t i a l l y , a series of broad responses were observed. The INDOR responses were optimized by slightly altering the power outputs of the external- and sweep- oscillators. In this instance, scanning a small range (- 0.3 Hz) of the monitoring frequency to insure the peak of the transition was being monitored, further helped to decrease the line-widths of the INDOR responses. Within ten minutes of adjustments the INDOR spectrum's resolution approximated that of the normal spectrum (see Fig. 2). In a similar manner, other lines corresponding to anomeric resonances were monitored. Sucrose octa-acetate(2) A sample of 2_ (m.p. 98-100°) was kindly provided by Dr. J. F. Manville and was used as a solution of ca. 20% w/v in acetone-d,. — o D-Ribose(3) The D-ribose (m.p. 98-99°), (Pfanstiehl Labs.), was lyophilized sev-eral times with D„0, then made to c_a. 30% w/v solution with ca_. 30 mg of sodiumdiraethylsilapentane sulphonate (DSS) added as internal standard. B. 1H-[ 1 3C] INDOR EXPERIMENTS Methyl 2,3,4,6-tetra-0-acetyl-a-D-glucopyranoside(4), m.p. 102° l i t . 6 101°. Methyl 2,3,4,6-tetra-0-acetyl-g-jj-glucopyranoside(5), m.p. 105°; l i t . 6 104°. Methyl 2,3,4,-tri-0-acetyl-a-Q-xylopyranoside(6), m.p. 88-89°. Methyl 2,3,4-tri-0-acetyl-B-D-xylopyranoside(7), m.p. 116-117°. • * Methyl-a-D-glucopyranoside (8), m.p. 160-161 . Methyl-g-D-glucopyranoside (9), m.p. 102-104°. *** Methyl-ot-D-xylopyranoside (10), m.p. 93 . Methyl B-D-xylopyranoside (11), m.p. 154-156°. 1,2,3,4,6-Penta-d-acetyl-B-D-galactopyranose(12), m.p. 142-143°; l i t . 7 142°. * * o 2-Acetamido-2-deoxy-D-glucopyranose (13), m.p. 215 (d). * * o 2-Acetamido-2-deoxy-D-mannopyranose (14) , m.p. 127-130 . * From Eastman Organic Chemicals, Rochester, N.Y. ** From Pfanstiehl Laboratories, Waukegan, I l l i n o i s . *** From Koch-Light Laboratories, Colnbrook, U.K. The above compounds were synthesised by standard literature tech-niques or purchased directly from suppliers. The n.m.r. spectra of derivatives 4_-7_ and 1_2 were measured using ca. 35% w/v solutions in chloroform-d, contained in a 5 mm O.D. sample tube; tetramethylsilane was used as the internal reference for the f i e l d -frequency lock. N.m.r. measurements on compounds 8^ 11_, and 13_ and 14_ were made with samples prepared as follows: the derivative was lyophylised several times with deuterium oxide and then a solution (ca. 40% w/v) was made with deuterium oxide. The sample was measured using a 5 mm O.D. tube, with tetramethylsilane contained in a capillary tube as the spectro- ,; meter lock signal. The apparatus used for these experiments involves components orig-38 inally designed for heteronuclear spin-decoupling experiments with the addition of a frequency-frequency lock. In general, the system described can be used to obtain indirect spectra of any heteronuclei provided the appropriate irradiation module adapter is available. To obtain the required frequency-frequency lock, several standard frequencies [34 MHz, 34 MHz, 32 MHz] were taken from a Hewlett Packard synthesiser driver (model 5110B) and added together using Hewlett Packard double balanced mixers (model 10514A). The resultant 100 MHz frequency was amplified to 7.0 volts (r.m.s.). The oscillator tube (V-103) was removed from a 100 MHz V-4311 radiofrequency unit and the 7.0 volt, 100 MHz signal was injected via a B.N.C. connector at J-108. 13 The C frequency was obtained by using a Barry Research di g i t a l programmer (model LSC 7B) to provide the linear sweep control for the Hewlett Packard frequency synthesiser (model 5105B). The power output of the frequency synthesiser was suitably attenuated with a Telonic Industries attenuator (model T.G.-9050). The INDOR spectra were recorded either on a Varian Associates Strip Chart Recorder (model G-14) or a Hewlett Packard-Moseley Recorder (model 7005B). The Y-axis drive was obtained from the external-recorder output of the HA-100 V-4311 unit. The X-axis of the charts was calibrated dir-ectly using the time base, event-marker-output of the Barry programmer. The reproductiveness of the chemical shifts, obtained directly from 1000 Hz sweep width charts and repeated over a period of several months, is - 2 Hz (- 0.08 ppm, for ^ C shi f t s ) . Higher precision could be obtained by measuring the resonance frequency of each transition at much narrower sweep widths or by direct calibration via manual sweeping of the fre-quency synthesiser. The sole limitation in the accuracy of shifts ob-tained by either of these latter methods is then governed by the line widths of the INDOR responses themselves. 13 The general procedure used during these C INDOR measurements was as follows: a frequency-frequency lock was established as described above. Using the frequency-sweep mode, the instrument was placed in field-frequency lock in the usual manner via a tetramethylsilane refer-ence. A normal proton n.m.r. spectrum of the compound was measured 13 under conventional conditions. The C satellites of the appropriate resonance were then recorded at increased gain. The sweep oscillator of the Varian recorder was set to the precise resonance frequency (-0.1 Hz) of the s a t e l l i t e which was to be monitored. Then the receiver out-put of the spectrometer was connected to the Y-axis of the external re-corder. With the f u l l power output of the synthesiser connected to the 13 probe via a double tuned adapter, the C resonance was found by sweep-ing the synthesiser over the appropriate 10,000 Hz range at a rate of 13 100 Hz/sec. (for C resonances, the appropriate range is 25,140,000 to 25,150,000 Hz). The broad resonance i n i t i a l l y detected in this way was subsequently measured on a narrower sweep width (1000 Hz) with a some-what slower sweep rate (2 Hz/sec.) and attenuation of the irradiating power (81 db). There are several ways the experimentally determined shifts may be quoted. I n i t i a l l y , because of the frequency-frequency lock, the data. can be tabulated in absolute frequency units directly from the charts. However, these values would only be significant for a particular machine observing at ca. 100,002,505 Hz. Hence, i t is more convenient to com-13 pare them on a ppm basis with a standard reference, namely, the C shift of tetramethylsilane. Several groups of workers have given form-ulae whereby the experimental frequency data may be converted such that the C shift is given in ppm from that of tetramethylsilane and at a magnetic f i e l d such that the protons of tetramethylsilane resonate at 29 30 39 precisely 100,000,000 Hz. ' ' There are two general formulae which 30 have been used: The f i r s t , applied by White and Olah is based on a 29 relationship f i r s t formulated by Paul and Grant. r - r i TMS 613 r = " • 1 0 + 6H L . 'TMS where r = frequency at which a particular proton is measured frequency of the corresponding ^^ C resonance and 6^  = the proton chemical shift (in ppm) of the lock proton from tetramethylsilane. Both groups of workers which had previously used this formula stated that the ratio of Tmie, = 3.9769331, for both internal and external TMS tetramethylsilane. Our value for differed slightly from these other authors. Repeated measurements in various solvents gave the re-sults shown in Table 3. The differences between these values is note-13 worthy and indicates that solvent effects may play a role in C shifts. The value for neat tetramethylsilane was used for the conversion of data for aqueous solutions and the value for tetramethylsilane in chloro-form was used for the data obtained in chloroform solutions. TABLE 3 Resonance frequencies for the H and C resonances of tetramethyl-silane(TMS). Solution '''H-Frequency (v0-MHz) •*"3C—F requency (fQ-MHz) r v o " f o Neat TMS 100.002505 25.145632 3 9769334 TMS:Acetone (7 : 3) 100.002505 25.145628 3 9769341 TMS:Chloroform (1 : 1) 100.002505 25.145639 3 9769323 The second formula, suggested by McFarlane^, takes the form: 108 X . Y = obs. TMS (10 + f - 1006) where X T M C, =' chemical shift relative to the protons of tetramethylsilane resonating at exactly 100,000,000 Hz. ^obs = r e s o n a n c e frequency actually recorded. 6= chemical shift in ppm relative to tetramethylsilane of the protons used for the lock signal. 10 + f = frequency at which, the lock signal is obtained. If tetramethylsilane is the lock signal, the formula simplifies to 10 8 X . „ obs .  TMS 10 + lock sideband frequency This equation converts the experimental frequency data to a basis where 13 the C frequency is given in Hertz from that of the protons of tetra-13 methylsilane resonating precisely at 100,000,000 Hz. The C shift for 13 the methoxyl group of these sugars can be quoted in ppm from the C resonance of tetramethylsilane by taking the difference, X^g - X^, 13 and dividing by the factor, 25.15 Hz/ppm, for C spectra. 13 Several C shifts were calculated using both formulae; identical values were obtained. In general, the f i r s t equation was used to cal-13 culate the C shifts presented in Tables 1 and 2. Sample calculation: For the case of methyl 2,3,4,6-tetra-0-acetyl-a-D-glucopyranoside(4) using the 2505 Hz upfield sideband of tetramethylsilane to provide the proton lock signal. i " ' 100,002,505 100,002,505 25,147,025" " 25,145,693 r - r c 6 = i TMS = . 10 TMS r T M S _ ... . 100,002,505 25,145,693 3.9767131 - 3.9769323 i n 6 r r n t-J-U- = - 55.12 ppm. 3.9769323 The negative sign indicates the shifts are to low f i e l d (higher frequency) of tetramethylsilane. C. OTHER HETERONUCLEAR' INDOR EXPERIMENTS By monitoring resonances in the proton spectra which exhibit the heteronuclear proton coupling, heteronuclear INDOR spectra were ob-tained for the following compounds: Trimethyl phosphite(15) , [Aldrich Chemicals, Milwaukee, Wis.] Trimethyltin chloride(16), [Alfa Inorganics, Beverly, Mass.] Bromosilane(17)* Tetramethyl lead(lS), [Peninsular Chemicals, Gainsville, Florida] •k Sample donated by Dr. R M c L e a n . Chemistry Department; U.B<C The n.m.r. spectra for compounds 15, 16 and 1_8 were measured using ca. 40% w/v solutions in chloroform—d, contained in a 5 mm O.D. sample tube. Compound 1_7 was measured as a 95% w/v solution containing 5% tetramethylsilane in a sealed pyrex tube. For a l l samples, tetramethyl-silane was used as the internal reference for the field-frequency lock. The appropriate irradiation module adapters were described else-38 where. In most, instances, one adapter could be used for several nuclei provided their resonating frequencies were sufficiently close. For 31 example, the P module adapter (40.481 MHz) could be used to irradiate both the 1 1 7 S n (35.626 MHz) and 1 1 9 S n (37.272 MHz) resonances. The chemical shifts for each heteronucleus are reported relative to tetramethylsilane resonating at precisely 100,000,000 Hz. The prob-+ able maximum error present in most of these measurements is - 5 Hz. SECTION II THE SYNTHESIS OF SOME NOVEL PHOSPHORUS AND ORGANOMETALLIC DERIVATIVES OF MONOSACCHARIDES. INTRODUCTION Although the f i r s t organometallic compounds were synthesized in the late nineteenth century, only in the last twenty years has there been a significant increase in the study of these carbon metal bonded substances. In fact, such studies have now tended to mediate di f f e r -ences, which previously had been stressed by most elementary chemistry textbooks, between organic and inorganic disciplines. It is the intention of this portion of the thesis to investigate the synthesis of carbohydrates containing carbon-metal bonds, speci-f i c a l l y C—P, C-Sn, C—Si and C—Pb bonds. The synthetic approach used in the preparation of these derivatives involves the formation of l i t h i o -organometallic reagents which are then used in nucleophilic displace-ment reactions on sugar tosylates and sugar epoxides. I n i t i a l l y , apart from the Grignard reaction and the synthetic use of organolithium reagents, concern with the aspect of organometallic chemistry remained in the domain of the inorganic chemist. However, with the development of such reagents as sodium borohydride, lithium 40 aluminum hydride and a variety of their derivatives as reducing agents , 41 in addition to hydroboration of olefins with boron hydrides , Wittig's 42 use of triphenyl phosphoranes for methylenation of carbonyl compounds 43 and cross coupling reactions between halides and copper alkyls , i.e. RX + R^CuLi R-R* + LiX + R'Cu there has been a more distinct emphasis on the synthetic organic ap-plications of organometallics. In the carbohydrate area, the f i r s t significant use of metals was 44 in Reeves' investigation of copper-ammonium complexes of free sugars. By studying spectroscopic absorption, optical rotation and conductivity changes taking place during the formation of these coloured complexes (see ), Reeves and his co-workers were able to make significant con-tributions to the area of carbohydrate conformations. As would be expected, an early start was made on reacting various Grignard reagents with sugar ketones and sugar epoxides. 6 3 These reac-tions provided a synthetic means for preparing branch chain sugars. More recently Co(CO)g has been used in the "oxo reaction" to synthesize specific branch chain sugars from unsaturated carbohydrates. The intermediate of this reaction involves the formation of a Co-C bond on the sugar. In a similar manner, methoxy-mercuration of unsaturated * sugars has produced Hg-C bonded monosaccharides which can readily be 46 isolated as the chloride salt. Upon hydrolysis these salts afford good yields of specific deoxy sugars. The importance of phosphorus in biological systems has prompted many chemists and biochemists to do extensive mechanistic and synthetic 49 studies related to this area. As an abundant element, this "metal" has some rather special properties arising from the fact that phos-phorus can readily exist in several stable valence states, namely: P(III) ; P(IV) and P(V). *199 Hg has a nuclear spin of 1/2 and natural'abundance of 16.84%. 47 1 Kreeyvoy and Schaeffer have examined the H n.m.r. spectra of methoxy 199 1 mercurated cyclohexane derivatives with respect to the vicinal Hg— H 3 coupling constants. The large values for the ^ ( t r a n s diaxial)=600 Hz compared to the range of ^ ( c i s ) = 100-200 Hz suggests these couplings are more sensitive to dihedral angle changes than vicinal 1 1 1 H— H couplings. Similarly, the H spectrum of methyl 3,4,6-tri-0-3 acetyl-2-deoxy-2-mercuric chloride-3-D-glucopyranoside gives a ^ ( g ) Hg(a) - 129 Hz. 4 8 In the carbohydrate area i t is well known that phosphate esters of monosaccharides are of considerable importance in the metabolic pathway.3^ Indeed, there are several classical methods for preparing these phos-phorylated sugar derivatives. Notably: reaction of phosphochloridates with alcohols 3^; ring opening of epoxides with dibenzyl hydrogen phos-52 phates ; and exchange of primary halides or of anomeric halides using 53 silver salts of phosphates. In contrast, carbohydrates containing carbon-phosphorus bonds have only recently been synthesized. One of the earliest methods of pre-paring these sugar derivatives involved the reaction of t r i a l k y l phos-* phite with a primary sugar halide, in the Michaelis-Arbusov manner , to 54 yield a phosphonate. More recently this method has been extended to the preparation of intermediates used in the synthesis of sugar anal-ogues with phosphorus as the ring heteroatom. 3 3 Good yields of sugar hydroxy-phosphonate derivatives have been re-ported for the reaction of dimethyl phosphite with sugar ketones in the presence of a strong base. 3 6 Although these derivatives tend to be some-what unstable, the analysis of their *H n.m.r. spectra has provided * The mechanism of the Michaelis-Arbusov reaction has un t i l recently been a subject of controversy. The pathway presently accepted for this reaction is as follows: ^0—Et (EtO) P-OEt + R-Cl — • (EtO) + CI + •(EtO)_P/^ + EtCl R where the formation of the phosphoryl bond is the driving force behind the reaction. information on the stereospecific relationship between vicinal P--C-C-H 57 ' -couplings and dihedral angles.' ' 58 Moffatt and co-workers , through the development of a'new stable Wittig reagent, [Ph^P^CHP(0)(OPh) ] ,were able to'prepare a, 3^unsaturated phosphonates of sugar nucleosides containing keto and aldehydo func-v tions. By reduction of the unsaturated linkage, they converted these compounds into analogs of natural phosphodiesters. . '.[• Another approach to the formation of carbon-phosphorus bonds has involved the photochemical addition of phosphines^'' or thiophosphines^ to unsaturated sugars. . . . Many of the above methods require carbohydrate precursors which are often only obtained via a several-step synthetic sequence. The aim of this study was the investigation of more general methods for pre-paring sugar carbon-phosphorus bonds. Since toluenesulphonic esters (tosyl) of sugars can be easily pre-pared, these compounds were considered attractive starting materials, for the synthesis of carbohydrates with carbon-phosphorus bonds. The following discussion w i l l show that sugar diphenylphosphine oxides can be successfully prepared by using lithium diphenylphosphine reagent as a nucleophile to displace sugar tosylates or to open sugar * * This author recognizes that the decision to use tosylated sugar derivatives could limit the reactivity of the monosaccharide towards the nucleophilic reagent. For instance, sugar halides are known to.be. generally more reactive towards nucleophiles than these sulphonic esters. However, the object of this study was to investigate the most general methods for preparing sugar phosphonates; indeed, further studies w i l l l i k e l y show that the corresponding sugar halides w i l l react to give products in higher yields. epoxides. S i m i l a r l y , by preparing other lithio-metal reagents v i a stan-dard l i t e r a t u r e methods one can also successfully synthesize sugars con-taining carbon-metal bonds of t i n , s i l i c o n and lead substrates. This to our knowledge i s the f i r s t time that such derivatives have been prepared. RESULTS AND DISCUSSION A. PREPARATION OF DIPHENYLPHOSPHINE REAGENT The intent of this study was the preparation of sugars containing carbon-phosphorus bonds from readily available precursors. A f i r s t objective was to find a good nucleophilic reagent. I n i t i a l l y , attempts were made to form a Grignard reagent by reacting a diaryl chlorophosphonate with magnesium chips in an analogous manner to the Grignard alkyl halide reaction with magnesium.6''' This approach, however, proved somewhat unfruitful as materials remained un-reacted after several hours of refluxing. The entrainment method of forming Grignards was then considered. A search of the literature revealed that various substituted phosphine 6 2 oxides have been prepared by this technique. Generally, the reaction proceeds by the following steps: RX + Mg —fup • ^ CI) 3RMgX + (EtO)2PH • R2$MgX + 2EtOMgX + RH (2) 0 0 H2° ^ 2 R0PH .+ MgXOH (3) R-pMgX 0 R , B r > R2PR' + MgXBr (4) Depending upon the Grignard used and groups displaced, the yields of the substituted phosphine oxides varied considerably. Preparation of diphenylphosphine oxide magnesium bromide from phenyl magnesium bromide and dimethyl phosphite and reaction with 1,2: 3,5-di-0-methylene-6-0-tosyl-ct-D-glucofuranose(20a) led to a very small amount of O-tosyl displacement. This product, however, could not be isolated to determine i t s structure. Suspecting that the 0-tosylate function may not be a good enough leaving group, this Grignard reagent was also reacted with a sample of 2,3,4,6-tetra-0_-acetyl-a-D-glucopyra-nosyl bromide(31). Subsequent workup showed the major product, namely phenyl 2,3,4,6-tetra-0_-acetyl-a-D-glucopyranose, to be identical to 63 that obtained by Bonner from the reaction of phenyl magnesium bromide with the same halo sugar. The formation of the phenylated, instead of the diphenylphosphine oxide, sugar derivative can probably be explained 64 in terms of some recent work published by Hays. From kinetic experi-ments he found that i f any magnesium halide is present during the for-mation of diphenylphosphine oxide magnesium halide, the reaction is re-* tarded. Whether the magnesium halide restricts the reaction by complex formation is presently under investigation. 6^ If retardation is a fac-tor in the above reaction, some phenyl magnesium halide, a very reactive species, w i l l s t i l l be present and the glucosyl halide can react with i t to produce the phenylated sugar glycoside. * Under normal conditions, this reaction w i l l not go to completion. Somewhat frustrated by the lack of success in the Grignard area, attention was turned to the use of a l k a l i metals to generate nucleo-philes. Organosodium compounds have often been used in the preparation of 40 organic derivatives. However, the ionic character of these compounds often makes them insoluble in most organic solvents; hence, organo-metallic-sodium reagents were considered unsuitable for this study. Lithium, on the other hand, is one of the few a l k a l i metals to form 66 organometallic compounds with properties typical of covalent substances. As such, organolithium compounds tend to be very soluble in non-polar solvents such as tetrahydrofuran. The use of alkyl lithium reagents is not new to the carbohydrate 6 7 f i e l d . Overend and co-workers used various organolithium reagents to prepare branched chain deoxy sugars from sugar epoxides. In the case of methyl lithium reacting with methyl 2,3-anhydro-4,6-0-benzylidene-ct-68 D-allopyranoside, a methylated 1,2-unsaturated compound was formed. * Several methods for the preparation of lithium diphenylphosphine(19) have been reported in the l i t e r a t u r e ^ 2 The lithium d phenylphosphine adduct was chosen as the organo-metallic reagent in these displacement reactions for two reasons: (1) From previous reports ' i t appeared to be a relatively . stable compound at 0° C and could be prepared from readily available and stable precursors. (2) Due to the presence of phenyl groups, the proton n.m.r. spec-trum could be diagnostic in analyzing the product of the reaction with-out interfering with the rest of the sugar proton resonances. [A] Reaction of diphenylphosphine with n-butyl lithium. 02PH + LiC 4H g fr> 0 2PLi + C 4H 1 Q* The main drawback here involves the handling of the diphenyl-phosphine which is both toxic and d i f f i c u l t to work with. For these reasons, this method for synthesizing lithium diphenylphosphine was not considered further. 70 71 [B] Cleavage of triphenylphosphine with lithium. ' 03P + 2Li 0 2PLi + 0Li 71 Aguiar and co-workers prepared lithium diphenylphosphine via this technique and reacted i t with benzyl chloride and dichlorohalides. 72 [C] Reaction of chlorodiphenylphosphine with lithium. 02PC1 + 2Li > 0 2PLi + LiCl 72 Tamborski prepared lithium diphenylphosphine and an ana-logous series of group IV organometallic reagents by this method. In the following discussion, only B and C are considered as attractive methods for the synthesis of the lithium diphenylphosphine(19). The symbol 0 w i l l be used throughout this thesis to "represent a phenyl group. * * Historically, anhydrous tetrahydrofuran has been considered an excellent solvent for organolithium- reactions. Due to both the high solvation of lithium diphenylphosphine and the solu b i l i t y of the sugar tosylates in dry tetrahydrofuran, this solvent was used exclusively for a l l displacement reactions. Preparation of the nucleophilic reagent 19_ using Method B was at-tempted several times but generally proved unsuccessful. Following 71 Aguiar's procedure, the reaction appeared i n i t i a l l y to be proceeding well (solution became a scarlet colour). However, upon the addition of 1,2:3,5-di-0-methylene-6-0-tosyl-a-jj-glucofuranose(20a) l i t t l e or no displacement occurred. Other primary sugar tosylates si m i l a r i l y proved unreactive. The poor reactivity of this lithium diphenylphosphine reagent may arise from the presence of phenyl lithium, formed as a by-* 71 product in Method B. Aguiar recognized the desirability of selec-tively eliminating the phenyl lithium from this reagent. On the basis that phenyl lithium was expected to be a stronger base than diphenyl-phosphine, he suggested that t-butyl chloride be added to the reaction mass in order to remove this unwanted by-product. However, since i t is likely that some of the t-butyl chloride may also react with the lithium diphenylphosphine to deactivate the reagent, the addition of this alkyl halide was not attempted here. , * For some reactions carried out using Method B, few precautions were taken to eliminate any phenyl lithium formed. No reaction occurred upon the addition of the sugar tosylates. This suggests that l i t t l e or no free phenyl lithium could be present. If free phenyl lithium was pre-sent in solution, displacement of the O-tosyl group by the phenyl nucleo-phile might be expected. ** . Diphenylphosphine r e a c t s ^ i t h phenyl lithium to produce lithium diphenylphosphine and benzene. Another reason for non-reactivity could arise from the reaction temperature. Normally, at room temperature, the Wurtz coupling reaction can interfere with the formation of the lithium diphenylphosphine re-agent. Such a situation would lead to the formation of dimers as follows: 0 3P + 2Li > L i P 0 2 + Li0 LiP 0 2 + 0 3P 0 2PP0 2 + Li0 At higher temperatures, dimer formation may even predominate. The pre-sence of phenyl lithium seemed a drastic limitation to Method B. Thus, instead of considering the effects of lowering the reaction temperature, * attention was focused on Method C. The use of chlorodiphenylphosphine and lithium proved successful for preparing lithium diphenylphosphine(19) provided the reaction temper-ature remained below -20° C. To determine the reactivity of 19 as a nucleophile, a sample of 1,2:3,5-di-0-methylene-6-0-tosyl-a-D-gluco-furanose(20a) was used as a standard. 1_9_ was prepared at room temper-ature, 0° C, and -20° C following identical procedures. The results of the reactions on the sugar tosylate 20a are shown in Table 4. The amount of 19_ required to complete the reaction was taken as the point where the addition of one more ml aliquot of reagent caused the reaction * Upon the completion of this work, a communication was published ^ which uses Method B to prepare unsymmetrical phosphines in good yields. However, other workers in this department have had d i f f i c u l t y in pre-paring an active reagent via Method B. TABLE 4 Results of adding lithium diphenylphosphine(19) solution to 1,2:3,5-di-0-methylene-6-0-tosyl-a-D-glucofuranose(20a). Preparation temperature Amounts of Reactants sugar tosylate L i P 0 2 solution room temp. (22° C) 1.0 g 25 ml 0° C 1.0 g 20 ml -20° C 1.0 g 12 ml mixture to turn and remain the characteristic crimson colour of the lithium diphenylphosphine solution. On this basis, Table 4 shows that the reagent prepared at -20° C is more "reactive" than those prepared at higher temperatures since less lithium diphenylphosphine reagent is required to react with 1.0 g of sugar tosylate. Preparation of the re-agent at temperatures much lower than -20° C required more extensive reaction times with l i t t l e improved reactivity. Normally, the lithium diphenylphosphine reagent turned from crimson to pale yellow during the addition to sugar tosylates. The usual procedure was to add reagent to the reaction mixture until the crimson colour remained, then to add ca. 10% excess reagent. Thus, i t proved a simple matter to determine i f a sufficient amount of 1_9_ had been added to the reaction vessel. In the case of slower reacting sugars (i.e. secondary tosylates) the reaction vessel was usually "topped up" with lithium diphenylphosphine reagent the following day. B. REACTION OF PRIMARY 0-TOSYLATES Other studies have shown that most primary sugar 0-tosyl substituents 77 are good leaving groups in nucleophilic displacement reactions. A series of these sugar derivatives were prepared by standard literature techniques and reacted with 19 in tetrahydrofuran at 0° C;. the reactions were performed under nitrogen atmosphere to insure the absence of both oxygen and water vapour. The amounts of reagent used and the resulting yields of product are presented in Table 5. I n i t i a l l y , there was some concern whether the products were sugar 78 phosphines or phosphine oxides; Infrared spectroscopy proved somewhat ambiguous here even though the ^ stretch frequency is characteristic at-1200-1350 cm *. Overlapping stretching frequencies from the bonds in the sugar moiety made unequivocal identification in this phosphoryl-stretch region impossible. To determine the valence state of phosphorus, a sample of 20b was dissolved in a small amount of acetone, containing 3% hydrogen peroxide, and l e f t at room temperature for several days. If 20b was a phosphine instead of the expected phosphine oxide, this oxidation procedure would change i t s physical properties. Workup of this reaction mixture gave a product identical in every nature to the starting material, thus confirming the i n i t i a l formation of the phos-phine oxide. This result is not surprising since the relatively harsh workup procedures for these displacement reactions should immediately oxidize a l l the products to the phosphine oxide derivatives. One important and, at f i r s t , rather surprising result present in Table 5 is the facile reaction of 1,2:3,4-di-0-isopropylidene-6-0-tosyl-a-D-galactopyranose(22a) with 19_. This sugar tosylate has previously 79a 79b been shown to react only slowly with fluoride c, methoxide , and L i P 0 2 19 Flow Sheet 2 TABLE 5 Product yields obtained from the treatment of primary sugar tosylates with lithium diphenylphosphine(19). REACTANTS PRODUCT % YIELD Sugar Grams 19_(ml)* 20a 2.0 20 20b 62.2 21a 2.0 40 21b 59.1 22a 2.0 15 22b 74.5 23a 2.0 25 23b 84.6 * If the reaction of lithium with diphenylchlorophosphine afforded a quantitative yield there would be 0.139 g of 19 per ml of reagent. iodide nucleophiles. However, i t reacts readily with such nucleo-p h i l i c reagents as potassium thioacetate and sodium azide in dimethyl 80 81 formamide to give the corresponding 6-thioacetyl and 6-azido derivatives. It has been suggested that the success of displacement reactions on 1,2:3,4-di-0-isopropylidene-6-0-tosyl-ct-D-galactopyranoside(22a) depends 82 upon the charge character of the nucleophile. The electronic effects of both the ring oxygen and the axial C—4 oxygen tend to prevent nega-tively charged anions such as fluorine, iodine or methoxide from dis-placing the tosyl group of 22a in an S^ 2 manner. On the other hand, more '.'neutral" species, such as thioacetate and azide w i l l tend to re-verse the polarity of the new bond in the transition state, resulting in an attractive interaction (lower energy). The lithium diphenyl-phosphine nucleophile f a l l s into this latter category, possibly due to the presence of the lone pair on the phosphorus atom. The structures of the products listed in Table 5 were determined from their *H n.m.r. spectra. In addition, compound 20b gave a satisfactory elemental microanalysis while the other products had close to the cal-* culated elemental percentages (see Experimental). * The compounds giving unsatisfactory analysis were syrups. Exposure of these substances to air at normal room temperatures over a period of two months caused slow decomposition. It is likely that a strongly acidic impurity, such as chlorodiphenylphosphine oxide, is present which readily degrades the carbohydrate over a period of time. Consider, for example, the 1H n.m.r. spectrum of 22b shown in Fig. 13A. A l l the major lines in the spectrum can be assigned to specific protons of the molecule. I n i t i a l l y , comparison of this spectrum with that of 22a helped to identify some of the resonances. Further information for 31 1 this assignment came from the P decoupled H n.m.r. spectrum (Fig. 13B). By comparing Figs. 13A and 13B, one can determine the values for a l l of 31 1 the P— H coupling constants. It is also noteworthy that the H-6 protons, which exhibit relatively large couplings to phosphorus, are both shifted to much higher f i e l d than normally expected. 31 1 Knowing which spectral lines are the result of P— H couplings allows one to perform heteronuclear INDOR experiments (see Experimental 31 of Section 1) to obtain the P chemical shift of 22b. Since several non-equivalent protons are coupled to phosphorus, two lines of the proton 31 * spectrum must be monitored to obtain the complete P INDOR spectrum. 31 One line must be associated with the P state, while the other must a 31 be associated with the P Q state. p 31 Figs. 14A and 14B show the P. INDOR responses of 22b obtained from 31 monitoring respectively an a and a 6 P-coupled transition of the H-62 resonance. The spectra are broadened by couplings from the phenyl protons in addition to the couplings from the sugar ring protons. The If the protons coupling into phosphorus are equivalent, the energy level system approximates that of an AX system (see Appendix A). Moni-toring an A transition w i l l give INDOR responses for a l l the X tran-sitions. This situation exists for the P INDOR spectrum of trimethyl phosphite as shown in Fig. 8. On the other hand, most of the deriva-tives discussed above approximate AMX systems (see Appendix B) and at least two lines must be monitored in the proton spectrum to obtain INDOR responses for a l l the X transitions. FIG. 13. (A), P a r t i a l 100 MHz H n.m.r. spectrum of 22b_ i n chloroform-d 31 s o l u t i o n . (B), Spectrum with the i r r a d i a t i o n of the P frequency (40.482 000 MHz) using noise modulation (500 Hz). 40,482,924 Hz FIG. 14. (A), P INDOR spectrum o f 22b i n c h l o r o f o r m - d s o l u t i o n ; o b t a i n e d from m o n i t o r i n g an a t r a n s i t i o n o f the H-6 2 resonance. ( B ) , 3 1 P INDOR spectrum o b t a i n e d by m o n i t o r i n g a B t r a n s i t i o n o f the H-6- resonance (see t e x t ) . P chemical s h i f t i s the mean value of the sum of these two INDOR spectra. 1 31 The H— and P— n.m.r. data f o r the primary sugar phosphine oxide derivatives are given i n Tables 6 and 7. A l l assignments of chemical s h i f t s and coupling constants were checked using a LA0C00N III computer * 31 programme. Extensive use was also made of P decoupling c a p a b i l i t i e s to assign resonances i n these spectra. Some general trends can be recognized from the data presented i n these two Tables. F i r s t , as mentioned previously, the chemical s h i f t s of the H-6 protons (H—5 protons i n the case of 23b) are a l l to much higher f i e l d (- 2 ppm higher) than normally found i n sugar d e r i v a t i v e s . This r e s u l t s from the sh i e l d i n g e f f e c t of the phosphorus moiety bonded d i r e c t l y to the carbon on which these hydrogens reside. As w i l l be shown further on, t h i s u p f i e l d s h i f t can be used to determine the p o s i -t i o n of various organometallic substituents on monosaccharide ri n g s . Here, t h i s s h i f t provides a diagnostic means f o r confirming that the primary 0_-tosyl group has been displaced by the diphenylphosphine anion. The r e s u l t i n g product i s subsequently oxidized to the phosphine oxide 31 d e r i v a t i v e during workup. Secondly, the P s h i f t s , which were measured i n d i r e c t l y from the *H n.m.r. spectrum using the heteronuclear INDOR * O r i g i n a l l y provided by Dr. A. A. Bothner-By, Chemistry Department, Mellon I n s t i t u t e , Pittsburgh, Pa., but further modified by Mr. R. B. Malcolm (Chemistry Department, U n i v e r s i t y of B. C.) f o r use on an I.B.M. 360/67 computer. Table 6. Chemical s h i f t s (T) f o r primary diphenylphosphine oxide monosaccharides. Compound 31p. 31p# H-l H-2 11-3 H-4 H-5 H-6 H-62 0 OMe 2 0 fo+ 40,481,879 85.4 4.08 5.62 5.70 5.92 5.55 7.27 2.2-2.8 -R=H 5.05( s) 5.28(q) t 21 b i 40,481,940 83.9 5.40 5.42 S.22 5.23 • 4.63 4.60 5.06 5.03 5.55 ~ \ 5.61 7.38 7.58 7.17 7.43 2.2-2.8 2.0-2.6 6.86 6.70 : 40,481,910 84.7 4.65 5.80 5.48 • 5.75 5.69 7.21 7.39 2.2-2.6 -*= C H3 8.5-8.8 2 3 § f 40,481,971 ' 83.2 4.16 5.44 a. "-S.7 ^5.7 -H-S1 H-52 7.18 2.2-2.6 OH 4.60 R=CH3 8.76 8.80 t i n CDClj solution. £ i n acetone - d^ solution. * Freo^ency i n H 2 r e l a t i v e to the protons of tetramethysilane resonating at exactly 100,000,000 Hz. # i n parts per m i l l i o n (ppm) from D^g» s e e footnote f or Table 8. 6 values taken d i r e c t l y from spectra without i t e r a t i v e computer analysis. TABLE 7. Coupl ing constants (Hz) f o r primary diphenylphosphine ox ide monosacchar ides. Compound J 1 . 2 J 2 . 3 J 3 , 4 J 4 . 5 5 , 6 ! J 5 , 6 2 J 5 , P \ , 6 2 J 6 2 , P 20b 3.8 <0.5 2.1 2.1 7.1 8 .9 - 12.4 . 3 . 8 10.1 9 . 3 9 .3 9 . 3 3.0 9.4 - 1 4 . 8 6 .8 15.1 21b 3.8 10.0 9 . 5 9 .5 9.4 3.1 9 . 3 - 1 5 . 0 6 .8 14.9 22?b 4.9 2.5 7.8 1.8 7.3 5.5 10.8 - 1 5 . 4 12.7 9 .7 23tt 3.7 <0.5 - -J 4 , 5 l 6.5 J 4 , 5 2 - 5 1 , 2 J 5 P 1. J 5 P 2, 9 .5 t i n CDCl j s o l u t i o n $ i n acetone - d^ s o l u t i o n 6 values taken d i r e c t l y from s p e c t r a without i t e r a t i v e computer a n a l y s i s . technique, a l l f a l l within a narrow range. The shifts are quoted in both absolute frequency units from the protons of tetramethylsilane (TMS) resonating at precisely 100,000,000 Hz, and in parts per million (ppm) from P^O^-These shifts can be shown to be indicative of phosphine oxide der-ivatives. For instance, n.m.r. data are presented in Table 8 for a series of diphenyl phosphorus derivatives in different phosphorus va-lence states. Both the phosphine and phosphine oxide of this series have environments similar to the corresponding substituted carbohydrate derivatives which could arise from the reaction of 19 with a primary sugar tosylate. The phosphine resonance is ca_. 2000 Hz (45 ppm) to : lower frequency (higher field) of that of the phosphine oxide. However, 31 the value of the P shift for ethyl diphenylphosphine oxide is very close to the values for the phosphorus sugar derivatives lis t e d in Table 6 and thus provides further evidence that these sugars are diphenylphosphine oxide derivatives. Table 7 also shows that for compounds 21b and 22b, one of the ^ couplings is much smaller than the other. By analogy with normal carbo-hydrate systems this would suggest that the favoured C—5-C—6 rotamer for these two derivatives to be as shown in |FJ , where 180° and 60° dihedral angles exist between vicinal hydrogens. This assignment is based on Table 8. N.m.r. data for ethyl diphenyl-phosphorus derivatives in various valence states. COMPOUND* 31 P shift 2 3 3 * Hz ** ppm JP,H P,H JH,H 0 2§CH 2CH 3 40,482,131 79.2 11.3 17.3 7.6 + CH CH X C H 3 40,481,814 87.1 13.7 20.2 7.6 02PCH2CH3 40,480,269 125.2 <0.5 16.8 7.5 P(OMe)38 40,487,449 -27.1 10.6 * Frequency relative to the protons of tetramethylsilane resonating at precisely 100,000,000 Hz. ** In. parts per million (ppml^from P^^J The conversion to ppm was accomplished by measuring the P shift of a sample of P(0Me)3 [Aldrich Chemicals, Milwaukee, Wis.] in two different ways. F i r s t , by direct means using P^O^ as the reference lock, P(0Me)3 gave a shift of -27.1 ppm from P 0^.' Secondly, via indirect means using the INDOR technique while monitoring a H n.m.r. transition. Other compounds measured via INDOR were converted to ppm relative to P^0^ by taking the frequency difference between P(0Me) and a specific compound and dividing by the factor 40.481 Hz/ppm. Note that higher frequency means lower f i e l d in these experiments. # These compounds were kindly provided by Mr. I. Armitage of the Chemistry Department, University of B. C. considerations of non-eclipsed rotamefs only. However, for the par-ticular compounds under discussion one cannot be certain as to what effect the diphenylphosphine oxide group may have on the magnitude of 1 1 * these vicinal H— H couplings. As w i l l be shown further on in this 3 discussion, the values for Jp ^ can also be used to assign rotamer populations. Due to the equivalence of H-6 and H—5 proton resonances in com-pounds 20b and 23b respectively, l i t t l e information concerning the rota-mer populations of these derivatives can be derived from the vicinal *H—*H coupling constants. * 83 A recent review article has shown that electronegativity effects can greatly alter expected vicinal couplings in carbohydrates. At this stage i t is convenient to make some specific comments con-cerning some of the *H n.m.r. parameters lis t e d in Tables 6 and 7. The ring *H—*H coupling constants for compounds 20b-23b can be used to assign conformations. As expected, the large values for 3 a n c* J4 5 4 * support the form as being the favoured conformation for 21b. Such, 84 1 however, is not the case for 22b. Cone and Hough have shown from H n.m.r. studies of some di-O-isopropylidene-galactopyranose derivatives that the cis fusion of two isopropylidene rings forces the sugar to adopt a non-chair conformation. The coupling constants reported for these derivatives, one of which was 22a, are typically: 2 ~ 5.0 Hz; J 2 3 = 2.4 Hz; J 3 4 = 8.0 Hz and 5 = 1.4 Hz. As shown in Table 7, the coupling constants for 22b are almost identical to these values. It is noteworthy that the value of J 3 4 for a l l these compounds immedi-ately precludes either chair form. The most lik e l y conformation satis-84 —— fying the n.m.r. data is the skew conformation (see \GJ). * 4 The designation Cj indicates that the compound is in the chair conformation with C-4 above and C—1 below the plane formed by the other four ring atoms. Usjng this nomenclature, the complementary chair form would be designated C^. ** 85 Hydrogen bonding studies have also shown that cyclohexane der-ivatives having, two cis fused isopropylidene rings cause the cyclohexane ring to adopt a skew conformation. Both compounds 20b and 23b have values of J_' = 0, which is char-acteristic of a dihedral angle of ca_. 90°. This indicates C—2 and/or C—3 are puckered'*out of the plane formed by the other ring carbon atoms. In the case of 20b, the similar magnitude of ^ a n d J3 4 allows one 3 3 86* to restrict the possible conformations to the T„ and V forms. * The letters T and V designate respectively twist and envelope con-formations, while the subscripts and superscripts are used to indicate which atoms are below or above the plane formed by the other ring atoms. C. REACTIONS OF SECONDARY TOSYLATES AND EPOXIDES Generally, nucleophilic displacements of secondary O-sulphonic 82 ester groups are d i f f i c u l t to achieve. This decreased reactivity at secondary positions has been attributed to both steric effects and to an unfavourable alignment of dipoles in the transition state of the reaction. For instance, 2-sulphonates of pyranosides are normally not displaced by nucleophiles. Their lack of reactivity has been attributed to the unfavourable dipole alignment in the transition state of both 82 the a and 8 anomers. On the other hand, steric effects from the pre-sence of "8-trans-axial" substit.uents with respect; to the sulphoxy group, impair displacements by benzoate ion in the a anomer of 1,2,4,6-tetra-O-benzoyl-3-O-tosyl-D-glucopyranose. However, the 3anomer does 87 undergo displacement by benzoate anion. 88 In a previous study, Hall and Steiner had shown that the O-mesyl (methanesulphonic) group of 3,6-anhydro-l,2-0-isopropylidene-5-0-mesyl-a-D-glucofurano5e (24a) can be readily displaced by a fluoride ion. Encouraged by this result, a sample of 24a was treated with a three-fold molar excess of 19_ to give, after subsequent workup and sublimation,, a 31% yield of diphenyl ^3,6-anhydro-l,2-0-isopropylidene-a-L-idofuranosej 5-C-phosphine oxide(24b) . The normal *H n.m.r. spectrum, together with the P decoupled spectrum of 24b are shown in Figs. 15A and 15B respec-tively, The similarity between the normal spectrum (Fig. 15A) and that 88 of 3,6-anhydro-5-deoxy-5-fluoro-1,2-0-isopropylidene-a-L-idofuranose helped to i n i t i a l l y assign the resonances. A l l lines in the spectrum can be assigned to specific resonances even though 24b gives an unsat-31 isfactory elemental analysis. The P decoupled spectrum (Fig. 15B) was particularly useful for determining the position of the phosphorus moiety. Refinement in the chemical shifts and coupling constants were accomplished via computer analysis using the LA0C00N III programme. These values are given in Tables 9 and 10. The coupling constants for the furanose ring of 24b are similar to. those found for other 3,6-anhydro-l,2-0-isopropylidene-L-idose deriv-88 3 atives. These values imply that the furanose ring favours a conformation. Furthermore, the small value of J. (. is indicative of 4, b the ido configuration. In contrast, the gluco precursor, 24a, has J. _ =4.4 Hz which is characteristic of 3,6-anhydro-glucose derivatives. " s b . • Beyond this point, the comparison between 24b and 3,6-anhydro-5-deoxy-5-f luoro-1,2-0_-isopropylidene-a-L-idofuranose breaks down; in the former compound, J r , and J r , are both large, while in the latter, these D.O, b ,o 1 2 two J values are both small. This suggests a conformational change in. the 3,6-anhydro ring. In order to obtain two large vicinal H^—^ H J12/ 4*3 b 1 62 H2 H 3 J6l,62A AJ6z'61 U U 4 ^,5 AM A -6,5 J4,P/ ft J5,6, 4.0 5.0 6.0 I , I T 7.0 FIG. 15. (A), P a r t i a l 100 MHz H n.m.r. spectrum of 24b i n benzene-d,-31 chloroform-d(4:l) s o l u t i o n . (B), Spectrum with P decoupling (40.482000 MHz) using noise modulation (500 Hz). Flow Sheet 3 TABLE 9. H and P chemical shifts (T) of secondary diphenyl phosphine oxide monosaccharides. COMPOUND 31 * 31p# H-l H-2 H-3 H-4 H-5 H-6j H-62 OMe OAc 4 -24^ 40, 481, 896 40, 481, 896 85.0 85.0 4.09 4.67 5.37 5.26 5.51 5.47 4.85 4.72 -6.6 6.82 5.93 6.06 - -2.2-3.0 2.1-3.1 26a 40, 481, 884 85.3 5.20 6.80 - - - - - 6.73 - 2.1-2.9 . 26b 40, 481, 887 85.3 5.19 6.56 4.72 »5.S 6.12 =5.5 5.72 6.80 7.98 2.0-2.9 28b 40, 481, 805 87.3 5.48 4.90 =6.5 - =6.3 - - 6.82 8.00 2.2-3.2 32 - - 5.44 5.17 6.45 4.64 4.97 5.62 5.86 6.70 8.0-8.1 2.1-2.8 30 40, 481, 794 87.6 5.46 5.06 6.74 4.84 -11-5 x 5.61 H-52 6.38 6.70 8.23 2.0-2.7 in CDClj solution t in cJiDg-COClj (4:1) solution * Frequency in Hz relative to the protons of tetramethylsilane resonating at exactly 100,000,000 II:. # in parts per million (ppm) from P«O f i ; see footnote for Table 8. TABLE 10. Coupling Constants (Hz) for secondary diphenyl phosphine oxide monosaccharides. COMPOUND J l , 2 J1,P J2,P J3,4 J3,P J4,5 J4,P J5,6 2 J5,P J 6 r 6 2 J 6 2 . P + 24b 3.8 3.7 -< 0.5 < 0.5 -3.5 3.6 -1.3 1.2 9.2 9.3 7.0 8.1 4.2 3.6 -9.0 14.6 11.6 <0.5 8.6 ~1.0 11.4 - - - - - - - - - ' + 26b <0.S 8.8 1.9 11.9 1.9 6.5 - - 4.2 - - -10.3 - -28b < 0.5 < 0.5 s 2.0 6.3 - - - - - - - - - -1.3 - 3 0 6.6 5.6 11.4 9.2 21.8 4.7 2.7 - -11.6 -30+ 3.4 c 0.5 5.4 8.0 5.6 9.1 J 3 , 5 2 1.2 8.8 2.6 \s2 3.0 -J 51, 52 -12.9 <0.5 V <0.5 + i n CDC13 solution # i n (f>D -CDC1 (4:1) solution couplings, the dihedral angles must be such that they l i e on either side of the Karplus curve. In other words, this implies dihedral angles of x° and (120 + x)° where x - 30-40°. Such a situation suggests the 3,6-anhydro ring favours the Sv conformation (as shown in |H | ) . This conformation would give dihedral angles between P—5 and the two H-6's of ca. 50° and 70°. Unfortunately, attempts to displace other secondary sulphonyl esters by the diphenylphosphine anion proved less successful. For instance, samples of methyl 4,6-0-benzylidene-2-deoxy-3-0_-mesyl-a-D-glucopyranoside and of 1,2:5,6-di-0-isopropylidene-3-0-tosyl-a-D-allofuranose were reacted separately with 19_ for several days. In each instance, the reaction product was found to consist of a multicomponent mixture, the * major substance being starting materials. During this period of the investigation, i t was f e l t that methods other than nucleophilic displacements of sulphonyl esters should be con-sidered for the preparation of secondary diphenylphosphine oxide deriv-89 atives. Epoxides, because of their relative ease of preparation , were considered as the most likely precursors for this purpose. It is well known that nucleophiles such as F", CI , OAc" and OH" react with sugar epoxides to open the epoxide bond. The reactivity of 19_ towards 2,3-anhydro-monosaccharides was used here to demonstrate the effectiveness of this synthetic approach for preparing secondary sugar diphenylphosphine oxide derivatives. 90 Previous studies have shown that methyl 2,3-anhydro-4,6-0_-benzy-lidene-ot-D-allopyranoside(25) reacts with nucleophiles to gave a trans diaxial opening of the epoxide. Furthermore, H n.m.r. spectroscopy is convenient for distinguishing between nucleophilic attack at C—2 or * In a related study, samples of me.th.yl 4,6-0-benzylidene-2-deoxy-a-D-gluco-hexopyran-3-ulose and 1,2:5,6-di-0-.isopropylidene-a-D-ribo-Eexofuran-3-ulose were separately treated with 19_ for several days. Monitoring of the reactions by t . l . c . showed a multitude of products. ** This trans diaxial preference is often referred to as the Furst-Plattner rule. at C—3. A product from the former reaction would have j = 3 Hz, while the latter reaction, i f i t occurred, would give a product having J 2 3 =10 Hz, since the H—2 and H—3 protons would then have a ca.180° dihedral angle. With this knowledge on hand, the allo-epoxide, 25, appeared to be an ideal model substrate. Reaction of 19_ with 2S_ gave a 65% yield of a product, 26a. Unfortunately, the configuration of this product could not be unequivocally determined from the *H n.m.r. spectrum (see Tables 9 and 10) because of the second order nature of the resonances. Acety-lation of 26a provided an almost quantitative yield of the corresponding 0-acetyl derivative 26b which gave a more suitable spectrum. The H—3 resonance was immediately identified by i t s shift to lower f i e l d upon acetylation. As mentioned above, a small coupling constant indicates a diequatorial arrangement between H—2 and H—3 and the value of J 2 3 = 1.9 Hz found for 26b confirms i t s structure to be as shown in Flow Sheet 3. In turn, this supports the structure of 26a, the i n i t i a l product of epoxide scission, as being diphenyl ^methyl 4,6-O-benzylidene-a-D-altropyranosidej 2-C-phosphine oxide. Further proof that the phosphorus moiety of 26a and 26b is attached to C—2 can be seen from the large upfield chemical shift of the H—2 31 resonance. The P chemical shifts for these and other secondary sugar phosphine oxide derivatives were again measured via the INDOR technique. 31 As is evident from the values lis t e d in Table 9, these P shifts f a l l in the expected range for phosphine oxides. Interestingly, f i r s t attempts to acetylate 26a using the acetic 91 anhydride-pyridine method proved unsuccessful. However, the use of 91 acetic anhydride-sodium acetate gave almost quantitative conversion to the acetate derivative. This observation is somewhat surprising since similar derivatives can be acetylated at C—3 using acetic anhydride-pyridine. It appears, therefore, that the presence of the diphenylphosphine oxide group causes an adverse effect on the reactivity of the C—3 position. The manner in which this effect is transmitted was not investigated further. The next substrate studied was methyl 2,3-anhydro-4,6-0_-benzylidene-q-D-mannopyranoside(27) which has been shown to give exclusive trans 90 diaxial opening when reacted with nucleophiles (i.e. nucleophile attacks at C—3). * The reaction of 19_ with 27_ afforded a 54% yield of 28a. Subse-quent acetylation of this product using acetic anhydride-sodium acetate gave an almost quantititative yield of an 0-acetyl derivative. The N.m.r. data are not report d f r this compoun  because of i t s poor solubility in most common organic solvents. small value of ^ (see Tables 9 and 10) strongly suggests a diequa-t o r i a l relationship between H—2 and H—3. This indicates that the struc-tures of 28a and 28b are as shown in Flow Sheet 3 and hence that 28a is diphenyl ^methyl 4,6-0-benzylidene-a-D-altropyranosidej 3-C-phosphine oxide. • , Both the alio-(25)and the manno-(27) epoxide derivatives discussed above are trans fused 4,6-0-benzylidene-hexopyranoside systems and, as such, exist in a r i g i d conformation. This r i g i d structure leads to exclusive trans-diaxial opening of the epoxide. Unfortunately, the *H n.m.r. data obtained from these compounds were limited because of the highly coupled nature of the H—4, H-5, and H-^ resonances in the spectra. One way of simplifying these spectra is to remove the 92 benzylidene group. Acid hydrolysis of 28a.using 1 N sulphuric acid, followed by normal acetylation gave a tri-O-acetylated product, 32. 1 The H n.m.r. spectrum of this product was greatly improved over that of 28a. Notably, H—3, H-4, H—5 and the H-6 protons were now sufficiently shifted away from one another that the spectrum was essentially f i r s t order. These shifts and the corresponding coupling constants were used to identify 32_ to be diphenyl ^methyl 2,4,6-tri-0-acetyl-a-D-altro-pyranosideg 3-C-phosphine oxide. As an extension of the reactions outlined above, i t was of interest to examine the product(s) arising from the opening of a non-rigid * epoxide. To make product identification easier, i t was desirable to choose a compound which could readily be examined by the n.m.r. technique. Treatment of methyl 4-0-acetyl-2,3-anhydro-B-L-ribopyranoside(29) with 1_9_ in the usual manner, followed by acetylation of the product, afforded a 48% yield of a white crystalline material, 30. Depending on whether the nucleophile attacks the epoxide at C—2 or C—3, there are * The removal of the 4,6-0-benzylidene group allows the pyranoside ring to react in either of two possible forms (see [T]) . As a result i t is often d i f f i c u l t to predict which product w i l l be formed. ** Sample provided by Mrs. L. Evelyn, Department of Chemistry, U.B.C. two possible products; furthermore, each product can exist in either of two chair conformations. N.m.r. data for 30 are given in Tables 9 and 10. The small values of J and J strongly suggest ca. 60° 4 , b l 4 , b 2 — dihedral angles between these protons. This eliminates the confor mations since one large value of J. would be required. The two re-maining possible chair structures are shown in [j] and |F|. Unfortunately the H— H coupling constants lis t e d in Table 10 do not provide unequivocal evidence about the structure of 30. The value of ^ = 5.3 Hz is not consistent with a trans diaxial coupling, thus suggesting structure jjl for 30. However, this would require that a l l substituents be in the axial position which conflicts with normal concepts of conformational analysis. The similar values of 3^ p a n d J4 p appear to provide l i t t l e addi-tional information towards resolving the structure of this reaction product. If the phosphorus substituent is attached to C—2 as shown in |K], one would expect p to be much smaller than the other couplings because of the long range nature of this coupling. Although this evi-dence supports structure [j] , the most convincing proof for this assign-ment comes from the chemical shift parameters listed in Table 9. Here, the H—3 resonance is to much higher f i e l d ( =2 ppm) than either H—2 or H-4. As shown previously, this upfield shift is indicative of a proton attached to the same carbon as phosphorus. One can only speculate as to why this conformation with a l l axial substituents would be favoured • ** over other forms. It is tempting to suggest here that the diphenyl-phosphine oxide group prefers to be eclipsed with the methoxy group on C—1. Unfortunately, time did not allow further experiments to be per-formed which may have provided more insight into why this molecule fa-vours the al l - a x i a l conformation. Previous studies have shown that upon treatment of 29_ with either Grignard reagent or sodium thiomethoxides, trans-diaxial opening of the epoxide occurs (i.e. attack at C—3). ** Boat and skew conformations were also considered. However, the H n.m.r. parameters did not provide as satisfying a f i t as in the a l l -axial chair conformation. So far, this study has shown how primary and secondary carbohy-drate diphenylphosphine oxide derivatives can be synthesized. However, as yet, no derivative containing the phosphorus group at the anomeric carbon (C—1) has been prepared. Here, displacement of sulphoxy groups at C—1 by 19_ is not a viable approach since generally, successful prep-* arations of an anomeric sulphonyloxy group are d i f f i c u l t . Reaction of 94 Brigl's anhydride , 3,4,6-tri-0_-acetyl-l,2-anhydro-a-D-glucopyranose, with 19_ could possibly give the desired anomeric phosphine oxide deriv-ative. However, the fact that the preparation of 1,2 anhydro-sugars requires a several step synthetic process precluded this compound from consideration here. On the other hand, 2,3,4,6-tetra-O-acetyl-a-D-glucopyranosyl bro-mide (31_) can be easily prepared and has been used successfully in dis-63 placement reactions with phenyl magnesium bromide. A sample of 31 ; was treated with ten times the molar amount of 19. Such a large excess is required because of the reactivity of the acetate substituents with 19. Subsequent workup and acetylation produced several products, the major one of which did not contain a diphenylphosphine oxide substi-tuent. By H^ n.m.r. spectroscopy, this product appeared to be * Attempts to prepare hexopyranosides with anomeric tosyl or mesyl substituentSg^ave only lead to the corresponding pyranosyl chloride derivatives. 1,2,3,4,6-penta-O-acetyl-ct-D-glucopyranose. This implies that 19 suc-* ceeds in only hydrolizing the bromide, 31. D. STEREOSPECIFIC DEPENDENCE OF 3 J 3 A comparison can be made between the. Jp ^ values for the com-pounds synthesized here and similar values reported from other n.m.r. studies. Recent work^'** 7 , 9^ with dimethyl hydroxy-phosphonate derivatives 31 1 has suggested that the angular dependence of vicinal P— H coupling ** constants follows a Karplus-type relationship. * From compounds having r i g i d conformations, such as 26b, 28b and to 3 some extent 30_ and 32, the Jp ^ values corresponding to a dihedral angle of = 60°, appear to f a l l into the range 6.6-11.5 Hz. These values are similar to those found in dimethyl hydroxy-phosphonate derivatives for 60° dihedral angles. At the time of writing this discussion, hindsight suggests that a more appropriate halo sugar to treat with lj^would be 2,3:5,6-di-0-isopropylidene-ct-D-mannofuranosyl chloride. The blocking groups used in this substance~do not react with 19. ** \ Recent evidence in^icajgs that^vicinal couplings between H and elements such as C, N, N and F depend on the respective dihe-dral angles between these two elements in a similar way as vicinal H— H couplings. The value of 21.8 Hz for p for compound 52_ should correspond t an - 180° dihedral angle. An alternate estimation of the coupling whi corresponds to an 180° angle can be obtained indirectly from the data presented for ethyl diphenylphosphine oxide (see Table 8). Since this acyclic compound rotates freely, the value of 17.3 Hz for J n „ is an P , H * average value. One can envisage, via Newman projection, three non-eclipsed stages in this rotation: J180° W J60° Assuming, for the present, that these are energetically the most favoured rotamers, the following equation can be formulated for the 3 jP H c o u P l i n g -* The fact that^the- methyl and methylene protons are equivalent and have a vicinal H— H coupling very similar to that of ethanol and other mono-substituted ethanes is sufficient proof for free rotation. •J 180° T °60° T u 60c average 3 (1) since J^QO = J'^n 0 n e r e > the equation simplifies to: j = J180° + 2 J60° " • average 3 (2) 3 Taking J^QO = 9.1 Hz as a representative value for Jp ^  (see Table 10) and substituting this together with J = 17.3 Hz into equation • - average n (2) gives: 17.3 = 3. therefore: J 1 0 f t 0 = 33.9 Hz l o l l J l g o o + 2(9.1) (3) 3 This value corresponds closely to the ( J „) previously found for 57 9 f i 8 0 ° dimethyl hydroxy-phosphonate derivatives. ' However, i t is 12 Hz greater than the value of p of 32, which should also have a dihedral angle of ca. 180°. In spite of this numerical discrepancy, i t is impor-tant to note that J 1 o r . o is greater than J ^ o and to this extent follows l o U o u 3 a Karplus-type curve. Furthermore, one can now use the J p ^  values to help determine the favoured rotamers of the primary diphenylphosphine oxide sugar derivatives given in Table 7. 1„ 1, As was previously shown, the H— H vicinal coupling constants for 21b and 22b suggested a favoured C-5-C-6 rotamer as shown in fiTT. H " \ I -4 O H However, there was some concern as to whether the diphenylphosphine oxide group could significantly affect the size of the coupling con-stants. The 3 J values of 20b, 21b, 22b and 23b f a l l in the range 9.3-10.8 Hz which is likely indicative of a dihedral angle of - 60° or 120°. These two possible values arise from the parabolic nature of the Karplus function. The 60° angle supports rotamer JL|. while the 120° angle suggests eclipsed rotamers as shown in IMI. The rotamers JMJ require the phosphine oxide groups to eclipse either * the ring oxygen or C—4 and normally one would intuitively discount them. However, these two eclipsed conformations correspond closely to the conformational situation which appears to pertain to 30_ where the methoxyl and diphenylphosphine oxide group have a 1,3 diaxial relation-ship; as a result, intuitive reasoning must be regarded with suspicion. * • — .—-—.—— The favoured conformation by propionaldehyde has been shown tp.be the form where the methyl group is eclipsed with the double bond. In the cases of compounds 20b and 23b, where the methylene protons 3 are equivalent, the Jp ^  values provided the basis for deciding which rotamers are favoured. Even with the rather sketchy stereospecific relationship between 3 Jp ^ and dihedral angles available at this time, i t was worthwhile returning to further consider 24b. Previous discussion indicated that 4 the 3,6-anhydro ring is in the TV conformation. From Table 10, J p (. * > \ and J ,. are 14.6 and 11.6 Hz respectively. These values f i t well into p,6 2 . A / the expected range for dihedral angles of 70 and 50° which are expected 4 for the T<- conformation. E. REACTION OF SOME LiM03 REAGENT WITH SUGARS Encouraged by the success of lithium diphenylphosphine as a nucleo-p h i l i c reagent for carbohydrates, attention was now turned towards the use of other similar organometallic species for these displacement reac-tions . Before starting this discussion, I should emphasize that my in-tent was to perform only a few preliminary experiments in order to deter mine the scope of these reactions. An exhaustive study was not intended Specifically, the reagents chosen were of the type LiM0 , where M = Sn, Si and Pb. The preparation of these reagents has already been 72 99 72 reported m the literature. ' Using the procedure of Tan\borski , the preparation of lithium triphenyltin(33) and subsequent reaction with the primary tosylates 20a and 22a gave respectively 36_ and 37 in good yields. The n.m.r. data l i s t e d in Tables 11 and 12 were used to assign structures to these compounds. Notably, the coupling constants and chemical shifts for these two products are almost identical to cor-responding diphenylphosphine oxide sugar derivatives. The only s i g n i f i -cant difference occurs in the chemical shift of the H-6 resonances. Here, the shifts are a further 1 ppm upfield from the derivatives having the phosphorus group on C—6 (see Fig. 16B and 16C). 20a R=Oj5,- 22a It was of interest to determine i f sugar epoxides would react as well with lithium triphenyltin as they had with lithium diphenylphosphine. Thus, 2 g of epoxide 29_were treated with an excess of 33. Normal workup procedures, followed by acetylation, gave a mixture of products. The mixture was chromatographed on a s i l i c a column to give a product, 38, in 21% yield. The 1H n.m.r. spectrum of this substance showed that two acetate groups and the correct number of aromatic protons were now present to indicate that epoxide scission had indeed taken place. A tentative assignment of structure could be made (see Experimental Section). However, this assignment requires the triphenyl t i n moiety, to significantly alter the normal chemical shifts of the ring protons which is certainly possible. • The other two lithio-organometallic reagents, 34_ and 35, were indi-vidually reacted with 20a to give in low yield, triphenyl ^l,2:3,5-di-0_-methylene-a-D-glucofuranosej 6-C-plumbane(39) and triphenyl ^1,2:3,5-di-O-methylene-a-D-glucofuranosej 6-C-silane(40). Again, *H n.m.r. spectroscopy proved invaluable in determining the structure of these products (see Tables 11 and 12 in Figs. 16D and 16E). The low yield of the triphenyllead sugar derivative, 39, i s not surprising. It is known that the longer bond lengths of higher atomic number metals result in lower bond s t r e n g t h . C o n s e q u e n t l y , the C-Pb o 3 6 R= Sn0, " " " " \ 3 9 R= P b 0 4 0 R= S i 0 c FLOW SHEET 4 bond is expected to be more labile than the C-P and C-Sn bonds. It is lik e l y that improved yields of these "lead" sugars could be obtained i f more care was taken during workup procedures (i.e. absence of strong acid or base). •, The very poor yield-of the triphenylsilane sugar-derivative, 40, was at f i r s t disappointing. Since the bond energies of C-C and C-Si bonds 1 0 0 are similar, one would expect that this product would be quite stable. Gilman and co-workers10''", however, have shown that lithium triphenylsilane has only moderate metalating a b i l i t y compared with other Group IVb metal t r i a r y l lithium reagents. This could be inter- ' preted as meaning that 35_ is not an effective nucleophile for. nucleo-p h i l i c displacements and hence that the low yield does not reflect any i n t r i n s i c i n s t a b i l i t y of the product. In this discussion, 20a has been used as a standard substrate for comparing the reactivity of-various lithium organometallic reagents towards the displacement of primary sulphonate esters. Fig. 16 shows a comparison between the n.m.r. spectra of .the starting material,20_a, and the products obtained from reactions with organometallic reagents. The corresponding coupling constants and chemical shifts for some of these derivatives are given in Tables 11 and 12. 5.0 6.0 70 T 8.0 CHjR 5.0 6.0 7.0 x ao F I G . 16. P a r t i a l H n.m.r. s p e c t r a of d e r i v a t i v e s of 1,2:3,5-di-O-isooropylidene-a-D-glucofuranose where (A), R= OTs; (B), R= P(0)0 (20b) ; (C), R= Sn0 3(36); (D), R= Pb0 3(39); and ( E ) , H= Si0,(50)in .chloroform-d s o l u t i o n . Table 11 Chemical shifts (T) for various sugar derivatives containing metallic substituents* Compound H-l H-2 H-3 H-4 H-5 H-6, H-62 R >; 0 36 + 4.07 4.54 5.66 5.86 5.81 5.51 6.16 ;5.6 8.01 8.25 R s 7 i § 5.50 2.3-3.0 37 f 6.01 ;S.8 7.98 8.20 R = CH3 8.6-8.9 R=H 5.35 5.53 2.4-3.0 39 + 4.17 5.73 5.87 j 6.45 "S.5 7.63 7.85 2.4-3.0 40 4.05 5.63 5.79 6.23 = 5.5' 8.13 8.35 R = H 5.17 5.40 2.4-2.9 * All values were taken directly from the spectrum without computer analysis, t In CDC1, solution. Table 12. Coupling constants (Hz) for various sugar derivatives containing metallic substituents. COMPOUND J l , 2 J2,3 J3,4 J4.S J5,6 2 \ , 6 2 3© . 3.8 <0.5 2.3 2.3 11.0. 6.8 -13.3 37 + 4.8 2-4 . 7.7 = 1.0 8.1 7.3 -12.7 3.6 = 1.0 = 2.0 = 2.0 8.5 7.8 -14.3 3.6 <1.0 2.2 2.0 10.6 ... -7.0 -12.2 A l l values were taken directly from the spectrum without computer analyses. t in C0C1_ solution. Notably, these five spectra are a l l similar in the low f i e l d region. However, with the substitution of various organometallic groups at C-6, the H-6 protons are a l l suddenly shifted to much higher f i e l d . As men-tioned previously, this shift can be taken as proof of the carbon atom to which the metal is bonded. An additional aspect of these derivatives is the presence of s a t e l l i t e lines in the proton spectrum due to the couplings between the spin 1/2 metal isotope and the H-6 protons (see * especially Fig. 16D). This section of the thesis has shown that lithium diphenylphosphine reacts well as a nucleophile in the displacement of primary O-tosylates of carbohydrates and in the opening of sugar epoxides. The products are diphenylphosphine oxide derivatives. The fact that the sugar pre-cursors used here are relatively simple to prepare and are by no means the most reactive derivatives available implies that this synthetic approach has considerable generality. Preliminary investigations indicate that lithium triaryl-group IVb metal reagents also exhibit good nucleo-p h i l i c character in reactions with some O-tosylate and epoxide sugars. The presence of metals with nuclear spin = 1/2 in these sugar deriv-atives would normally allow one to measure INDOR spectra of the metallic species. These measurements were i n i t i a l l y planned but not undertaken here for several reasons. F i r s t , the multitude of couplings between the metal and H-6 protons considerably reduces the intensities of their s a t e l l i t e lines in the 'H n.m.r. spectrum. Secondly, sol u b i l i t y of these derivatives in common solvents is limited and cannot compensate for the intensity loss due to the multitude of couplings. Future studies to overcome these two limitations w i l l be made when a wider variety of derivatives are available. Lithium triphenyltin especially shows good reactivity. At this point, i t should again be emphasized that these latter studies jwereby no means meant to be comprehensive in nature; the intent was only to show the potential- scope of these reactions. j One can foresee many possible extensions to the reactions discussed above. F i r s t , i t has been reported that dilithium phenylphosphine can 69 be easily synthesized. Reaction of this species with two equivalents of a primary-O-tosylated monosaccharide is l i k e l y to give a disaccharide with a phenylphosphine oxide bridge. Further extension of this theme could even lead to polysaccharides with C—phenylphosphine oxide link-ages. In addition, other lithium organometallic reagents could be used in an analogous manner to give C-,metal-linked polysaccharides. Such polymers may have some industrial significance since they offer the po-tential characteristics of organic compounds together with the conduc-tive properties of metals. It may also be of interest to examine the effects that these metal substituted sugar derivatives may have on bio-logical systems. The INDOR technique could prove quite useful for such a study. In a sli g h t l y different vein, one rather unexpected result arising from this synthetic study was the product obtained from the reaction of lithium diphenylphosphine(19) with methyl 2,3-anhydro-3-L-ribopyra-noside(29). Considering only chair conformations, there was a possibility of four different "products". Why the conformation with a l l the substituents in axial positions is favoured is a good topic for a future research proposal. This strange preference of the organo-metallic substituent to approach the axial position also seems to occur in the product formed from the reaction of lithium triphenyltin with 29_. It is obvious that these substituents must be placed on model compounds, such as a tert-butyl cyclohexane derivative, to try and determine the exact nature of these effects. Otherwise one can also make use of the numerous organolithium com-102 pounds which have been reported in the literature. These could pro-vide great potential in the synthesis of some rather novel carbohydrates. Ultimately, a study which would be most beneficial i s the develop-ment of methods to directly l i t h i a t e carbohydrates. Such species would be extremely useful for direct reaction with a number of organic, organo-•jnetallic and inorganic halides. A major obstacle to overcome in the preparation of these lithio-sugars would be the required presence of either a halide or highly acidic proton on the sugar, in order that li t h i a t i o n take place. At the outset of this thesis i t was emphasized that both the INDOR study and reaction of monosaccharides with organometallic reagents were related. The success of the INDOR experiments had a major influence towards encouraging the study of synthetic means for preparing sugars with carbon-metal bonds. In spite of the lack of s i l i c o n , t i n and lead INDOR data, the phosphorus INDOR experiment shows that heteronuclear INDOR can be useful in providing structural information concerning the products arising from these organometallic reactions. Notably, such a synthetic scheme would not have been undertaken without the desire to expand one's knowledge into other related f i e l d s . This writer strongly feels that more diversification of chemical studies into various related areas should be encouraged. The continuing proliferation of new instrumentation and the multitude of data present in the literature forces researchers to specialize. Being a "jack-of-all-trades" was frowned upon and s t i l l is to some extent. The comment is often made, and in some cases rightly so, that one can only be a master of one f i e l d and attempts to do research in several areas only leads to superficial knowledge in a l l fields. However, the concept of specialization i f taken r i g i d l y may significantly restrict the attain-ment of a research goal. Familiarity in many areas of science can often lead to the u t i l i z a t i o n of one's f u l l resources for investigating a specific area. This in fact, i s what research should be about; not just a series of individual investigations, but a conception of the relation-ships between various disciplines. It has been a continuing philosophy i n t h i s laboratory to encourage d i v e r s i f i c a t i o n i n order to prevent research from becoming too highly specialized and inward looking. This thesis i s the outcome of such an expanding interest i n various aspects of chemistry. EXPERIMENTAL General Methods Structural formulae associated with each compound are given flow sheets 2,3 and 4. Melting points were measured on a Fisher-Johns melting point apparatus and are uncorrected. ) Micro-analysis was carried out by Mr. P. Borda of the U.B.C. Chemistry Department. A l l *H n.m.r. spectra were measured on a Varian HA—100 spec-trometer operating in either the frequency-sweep or f i e l d -sweep modes using tetramethylsilane (TMS) as internal standard. Proton chemical shifts are a l l reported in the T scale. 31 (v) P N.m.r. spectra were measured using a Varian HA—100 spec-trometer operating in either of two modes: (a) The spectrometer was used in the locked f i e l d sweep mode with a capillary f i l l e d with I^O^ providing the lock standard frequency. (b) The spectrometer was set in the heteronuclear 31 INDOR mode (see Experimental of Section 1). The P spectrum was measured indirectly by monitoring an appropriate portion of the proton spectrum while scan-31 ning the P n.m.r. frequency range with a radio-frequency f i e l d set at low power. (i) ( i i ) ( i i i (iv) (vi) A l l infrared spectra were measured on a Perkin-Elmer 137 I.R. spectrometer using the nujol null method for solids. (vii) Optical rotations were measured on a Bendix ETL-NPL Automatic Polarimeter (type 143 A) using a 0.5 cm c e l l . A l l optical rotations were measured by Mr. Keith Wilson using chloro-form solutions. ( v i i i ) Nomenclature for sugar phosphine oxide derivatives followed 53 the precedent of L. D. Hall and co-workers. 31 (ix) P decoupling experiments were performed on a Varian HA—100 spectrometer. The decoupling frequency was derived from a Hewlett-Packard frequency synthesiser-driver (model 5110 B) and a Hewlett-Packard frequency synthesiser (model 5105 B). The double-tuned adapters and noise modulation equipment 38 have been described elsewhere. (x) Chemical shifts and coupling constants were checked using the LA0C00N III computer programme on an IBM 360/67. Preparation of Lithium Diphenylphosphine(19) The following procedures were used to prepare 19_ with the best results obtained from method (a). (a) 2.5 g of lithium metal (Matheson, Coleman and Bell, Norwood, Ohio) in the form of shavings were added to a two neck flask containing a magnetic s t i r r i n g bar and 40 ml of tetrahydrofuran (dried over sodium) under an atmosphere saturated with dry nitrogen. A septum cap was placed over one of the necks of the flask while the other was closed by a ground glass stopcock. The flask was cooled to ca. -20° C by slowly adding dry ice to an acetone bath. 8 g of chlorodiphenylphosphine (Aldrich Chemicals, Milwaukee, Wis.) in 10 ml of dry tetrahydrofuran were added, via the septum cap using a syringe, over a period of one : hour to the cooled stirred solution. The stopcock was opened occasi-onally to vent excess pressure from the system. The solution remained colourless until the addition was almost complete, then turned scarlet. In some instances this colour change, which signifies the formation of the lithio-diphenylphosphine complex(19) did not occur u n t i l after the addition of chlorodiphenylphosphine was completed. The solution was stirred at 0° C for a further four hours. The complex could be readily stored in a well-stoppered flask at 0° C for several days. If desired the excess lithium can be separated from the complex by f i l t r a t i o n under nitrogen atmosphere. Otherwise, a Syringe was used to remove the lithium diphenylphosphine solution from the flask. On the basis of quantitative yields, this procedure would produce 0.139 g/ml of lithium diphenylphosphine reagent. However, to compensate for the side reactions which occur during the preparation of this reag-ent, at least one equivalent excess of lithium diphenylphosphine(19) was added during the displacement reactions. Generally, the reagent could be prepared in up to 15 g lots without a significant decrease in yield. (b) Identical procedures were used as in (a) except the addition of the chlorodiphenylphosphine to the lithium metal was carried out at temperatures of (i) 0° C and ( i i ) 22° C (room temperature). See Results and Discussion for details of reactivity. (c) The same procedure was followed as in (a) except that only 0.5 g of lithium was used. This complex was found to be approximately half as active of a reagent for nucleophilic displacements as was (a). 71 (d) Using the procedure of Aguiar and co-workers , 3.6 g of t r i -phenylphosphine (Aldrich Chemicals, Milwaukee, Wis.) and 0.2 g of finely divided lithium metal were added to a stirred solution of 30 ml of anhydrous tetrahydrofuran under a dry nitrogen atmosphere. The solution was stirred for four hours during which time i t turned to a scarlet colour. Reactions of sugar tosylates with this reagent resulted in l i t t l e or no displacement. Attempted Preparation and Reaction of Grignard Phosphine Oxide: (a) 4 ml of phenyl dichlorophosphate (Aldrich Chemicals) were slowly added to 2 g of magnesium chips in anhydrous ether following normal Grignard procedures. The mixture was stirred and refluxed for several hours. Workup afforded only starting material. Similar results occurred using tetrahydrofuran as solvent. 64 (b) Following the procedure of Hays 7.0 g of bromobenzene and 1.5 g of magnesium chips in 30 ml of anhydrous tetrahydrofuran were re-acted in the usual Grignard manner. The mixture was refluxed for one hour followed by the slow addition of 1.8 g of dimethyl phosphite , (Aldrich Chemicals). The solution was refluxed a further four hours during which time the mixture turned a dark green colour. (i) 1.5 g of 20a in 10 ml of anhydrous tetrahydrofuran were slowly added to two equivalents of this Grignard reagent. The mixture was refluxed for 5 hours. Subse-quent workup in the usual manner and monitoring of the product by n.m.r. indicated that l i t t l e or no displace-ment had occurred (i.e. the methyl ''"H n.m.r. resonance of the tosyl group was s t i l l present in the correct amount for the reaction product to be essentially star-ting material). ( i i ) 1.0 g of 2,3,4,6-tetra-0_-acetyl-a-D-glucopyranosyl 103 bromide(31) was reacted with 15 equivalents of this Grignard reagent in a similar manner as (i) above. Workup and subsequent reacetyl'ation gave a product which was identified by n.m.r. to be 2,3,4,6-tetra-0-acetyl-8-D-glucopyranosyl benzene m.p. 156° C; l i t . 155° C. Diphenyl ^1,2:3,5-di-Q-methylene-a-D-glucofuranosej 6-C-phosphine oxide (20b) 104 2.0 g of lJ2:3,5-di-0-methylene-6-0-tosyl-a-D-glucofuranose(20a) m.p. 112° C, and 10 ml of dry tetrahydrofuran were placed in a two necked flask under dry nitrogen and sealed with a rubber septum cap. 20 ml (0.015 moles) of lithium diphenylphosphine solution(19) were slowly added with a syringe to the cooled flask ( 0 ° C). The scarlet colour of the lithium diphenylphosphine reagent was discharged immediately upon reacting with the sugar tosylate. Simultaneously, a heat of reaction was detected. The colour change was found to be beneficial in deter-mining i f an excess of 19_was present. A further 2.0 ml of lithium diphenylphosphine reagent was added after 1/2 hour to bring the reaction vessel back to the scarlet colour. The mixture was l e f t overnight at room temperature. The solution was then poured into ten times the volume of water, neutralizing with 1 N hydrochloric acid and stirred for several minutes whereupon a crude product precipitated. The crystals were f i l t e r e d and washed several times with water; crude yield, 1.8 g. Recrystallization from benzene: petroleum ether (30-60) afforded 1.3 g (62.2%) of 20b; m.p. 164-166° C. Found C, 61.53; H, 5.03; P, 8.00. Calculated f,,r C^H^C^P : C, 61.69; 75° H, 5.39; P, 8.23. [a]^ ^ +27° (c,1.07). Generally, these sugar phos-phine oxides were found to be very hygroscopic. From the relatively harsh workup procedures and the prolonged exposure of the product to ai r , one would expect the formation of a sugar phosphine oxide. Unfortunately, the phosphoryl stretch region in. 78 the infrared spectrum was obscured by other peaks. To prove the oxide was formed, a small sample of 20b was added to a solution of hydrogen peroxide (5%) in acetone, following the procedure of Mann and Millar 1 0**, to insure oxidation. Subsequent workup showed the product to be iden-t i c a l to 2_0_b_, thus providing satisfactory proof that the product of the reaction of sugar tosylates with lithium' diphenylphosphine was a sugar 31 phosphine oxide. Further proof was provided by the P n.m.r. chemical shifts (see Results and Discussion). 20a was reacted with several lots of lithium diphenylphosphine reagent in a similar manner as (a) but at various temperature ranges. The yields are given in Table 4 (see Results and Discussion). Although the elemental microanalysis for the following four com-+ pounds are not within the acceptable range (- 0.3%), the analysis data are presented to indicate the relative purity of these compounds. N.m.r. data, however, are.sufficient for determining the configuration and conformations of.the sugar phosphine oxide derivatives. Diphenyl ^methyl 2,3,4-tri-O-acetyl-ot-D-glucopyranosidej 6-C-phosphine  oxide(21b) 2.0 g (0.0056 moles) of methyl 6-0-tosyl-a-D-glucopyranoside(21a) and 10 ml of anhydrous tetrahydrofuran were placed under a nitrogen at-mosphere in a two necked flask. 40 ml (- 0.03 moles) of 19 were slowly added via a syringe to a reaction vessel cooled by an ice bath. The mix-ture was allowed to stand overnight at room temperature. The solvent was removed under reduced pressure followed by the addition of 15 ml of dry pyridine and 5 ml of acetic anhydride. The flask was shaken several times and then allowed to stand at room temperature for 12 hours. The solution was poured into ten times the volume of water, neutralized with I N hydrochloric acid and extracted-thrice with chloroform. The extracts were dried over sodium sulphate and evaporated at reduced pressure to give 2.5 g of crude white crystalline product. Recrystallization twice from aqueous ethanol afforded 1.7 g (59.1%) of 21b; m.p. 146-148°. Found C, 58.68; H, 5.59. Calculated for C 2 5H 2 g0 2P: C, 59.41; H, 5.74. Column chromatography of 21b did not improve the elemental microanalysis. Diphenyl p,2:3,4-di-0-isopropylidene-a-D-galactopyranosejj 6-C-phosphine  oxide(22b) 2.0 g (0.0048 moles) of l,2:3,4-di-0-isopropylidene-6-0-tosyl-a-D_ 107 galactopyranose(22a) in 15 ml of anhydrous tetrahydrofuran were treated in the usual manner with 15 ml (= 0.01 moles) of 19 and allowed to stand at room temperature for two days. During this time, a further 2 ml of 19_ were added to the solution in order to maintain. a scarlet colour in the reaction vessel. Addition of a large excess of water to the solution, followed by neutralization with 1 N hydrochloric acid, extraction thrice with chloroform and evaporation at reduced pressure gave 2.5 g of crude syrup. The syrup was separated on a s i l i c a gel column ( Malinkrot S i l i c a r CC 7) using 5% methanol in chloroform to afford 1.58 g (74.5%) yield of an amber syrup. *H N.m.r. identified this compound to be 22b. Found, 61.13; H, 5.94. Calculated for C24 H29°6 P : C ' 6 4 - 7 2 ' H> 6 * 5 2 -Diphenyl ^1,2-0-isopropylidene-a-D-xylofuranosej 5-C-phosphine oxide (23b) 2.0 g (0.0058 moles) of 1,2-0-isopropylidene-5-0-tosyl-a-D-xylo-* furanose(23a) was reacted in a similar manner as above with 25 ml (0.022 moles) of 1_9_. Subsequent workup gave 2.2 g of a crude syrup. Separation on a s i l i c a column using chloroform:methanol (9:1) afforded a 1.8 g (84.6%) of an amber syrup. The n.m.r. spectrum identified the product to be 23b. Found C, 60.99; H, 5.93. Calculated for C20 H23°5 P : C ' 6 2 A 0 > ti> 5 - 9 7 ' Sample kindly donated by Mrs. L. Evelyn, Chemistry Department, University of British Columbia. Diphenyl p,6-anhydro-l,2-0-isopropylidene-a-L-idofuranosej 5-C-phos-phine oxide(24b) 1.0 g (0.0036 moles) of 3,6-anhydro-l,2-0-isopropylidene-5-0-88 mesyl-ct-D-glucofuranose(24a) were reacted with 20 ml (= 0.014 moles) of 19 in a similar manner as described above. The reaction solution was allowed to remain at room temperature for two days. Workup in the usual manner afforded an amber syrup which was chromatographed on a s i l i c a column using methanol/chloroform (1:99) to give 0.475 g of a white cry-stalline material, m.p. 201-205° C. Sublimation at 120° (0.1 mm) afforded 0.420 g (30.2%) of 24b; m.p. = 208-209°. Found C, = 65.09; H, = 6.50; calculated for C^FL^C^P:. C, 63.5; H, 5.78. [ a ] 2 5 = +37° 31 1 (c, 0.95). Comparison of the normal and P decoupled H n.m.r. spectrum with related 3,6-anhydro-L-idose derivatives provided sufficient proof of structure (see Results and Discussion). Diphenyl ^methyl 4,6-0-benzylidene-a-p-altropyranosidej 2-C- phosphine  oxide(26a) 2.0 g (0.007 moles) of methyl 2,3-anhydro-4,6-0-benzylidene-a-D-108 allopyranoside(25) in 10 ml of anhydrous tetrahydrofuran were treated with 25 ml (- 0.02 moles) of 19_ as described above. The reaction ves-sel was immersed in an ice bath during the addition of 19. The solution was lef t overnight at room temperature. Ten times the volume of water was added and the solution was neutralized with 1 N hydrochloric acid, whereupon white crystals precipitated from the solution. Crude yield 2.8 g; Recrystallization from aqueous acetone afforded 2.4 g (68.5%) of 26a, m.p. 277-280°. Diphenyl pnethyl 3-0-acetyl-4,6-0-benzylidene-a-D-altropyranosidej 2-C- phosphine oxide(26b) (a) 1.0 g of 26a was acetylated using the acetic anhydride-sodium 91 acetate method. Workup in the normal manner followed by r e c r y s t a l l i -zation from aqueous ethanol gave an almost quantitative yield of 26b; m.p. 125-127°. A small sample was purified by column chromatography ( s i l i c a , 1% methanol in chloroform) for elemental analysis. Found C, 65.52: H, 5.75; P, 6.10. Calculated for C 2 8H 2 gO ?P: C, 66.01; H, 5.70; P, 6.29. [a]^ 5 = +68° (c, 1.13). 91 (b) Acetylation via the acetic anhydride-pyridine method only gave ca. 50% reaction (monitoring by t.l.c.) after two days. Diphenyl ^methyl 4,6-0-benzylidene-a-D-altropyranosidej 3-C-phosphine  oxide(28a) 3.5 g (0.014 moles) of methyl 2,3-anhydro-4,6-0-benzylidene-a-D-108 mannopyranoside(27) in 20 ml of dry tetrahydrofuran were treated in the usual manner with 35 ml 0.03 moles) of 19_. The solution was l e f t at room temperature overnight. The solution was treated in the usual mariner to give 3.9 g of crude material. Recrystallization from aqueous acetone afforded 3.6 g (53.7%) of 28a; m.p. 280° C. Found C, 67.0; H, 5.70; P, 6.84. Calculated for C^H^C^P: C, 66.81; H, 5.78; P, 6.85. Diphenyl ^methyl 2-0-acetyl-4,6-0-benzylidene-a-D-altropyranosidej 3-C- phosphine oxide(28b) (a) 1 g of 28a was acetylated via the acetic anhydride-sodium 91 acetate method to give an almost quanitative yield of 28b. Recry-stall i z a t i o n twice from aqueous ethanol afforded pure white needles, m.p. 223-225°. Found C, 65.74; H, 5.60; P, 6.57. Calculated for C28 H29°7 P : C ' 6 6 ' 0 1 ' H> 5 - 7 0 ' P> 6 ' 2 9 - [ a]p^ = +88° (c, 0.88). 91 (b) Acetylation by the acetic anhydride-pyridine method resulted in ca. 50% reaction (monitoring by t.l.c.) after three days. Diphenyl ^methyl 2,4-di-O-acetyl-g-L-xylopyranosidej 3-C-phosphine  oxide(30) 0.6 g (0.0032 moles) of methyl 2,3-anhydro-4-0-acetyl-3-L-ribopy-* ranoside(29) in 10 ml of dry tetrahydrofuran were treated in the usual manner with 15 ml (= 0.015 moles) of 19_. The reaction was le f t over-night at room temperature. The tetrahydrofuran was removed under vacu-um and the resulting white solids were acetylated using acetic anhy-dride and pyridine. After being allowed to stand for four hours, the * Sample provided by Mrs. L. Evelyn, Dept. of Chemistry, U.B.C. solution was added to ten times the excess of water and the mixture extracted thrice with chloroform. The extracts were evaporated at reduced pressure to afford white crystals. The product was recrystal-lized from ethanol/petroleum ether (30-60) to afford 0.65 g (48.1%) of 30_, m.p. 230-233°. A small sample was further purified on a s i l i c a gel column (1% meth-anol in chloroform) to provide a sample for elemental microanalysis. Found C, 60.63; H, 5.44. Calculated for C^H^OyP: C, 60.97; H, 5.64. m.p. 234-236°. [a]^ 5 = +63 (c, 1.39). Diphenyl ^methyl 2,4,6-tri-0-acetyl-a-D-altropyranosidej 3-C-phosphine  oxide(32) 100 mg of 28a were treated with 25 ml of 1 N sulphuric acid at 92 50° C for 4 hours. After standing overnight at room temperature, the reaction mixture was neutralized with barium hydroxide, f i l t e r e d and the precipitate washed twice with water. The water was removed under reduced pressure and the remaining white solid was acetylated using acetic anhydride-pyridine reagent. The mixture was allowed to stand for two days and then treated in the usual manner. The '''H n.m.r. spectrum indic-ated 90% removal of the benzylidene group with a corresponding increase in the acetyl methyl resonances to assign the product as 32_. Attempted Preparation of Some Secondary Diphenylphosphine Oxide Sugar  Derivatives (a) 1.0 g of methyl 4,6-benzylidene-2-deoxy-3-0-mesyl-a-D-arabino-109 hexopyranoside were treated with 15 ml of 19 reagent. The solution was left for several days at room temperature. Subsequent workup, as des-cribed above, gave a mixture which by t . l . c . appeared to contain at least six components. No attempts were made to resolve each component. (b) 1.25 g of l,2:5,6-di-0_-isopropylidene-3-0-tosyl-a-D-allo-* furanose were treated with 19_ in the same manner as described for (a) above. The product again consisted of a multi-component mixture. 103 (c) 2.0 g of 31_ were treated with 40 ml of 1_9. An ice bath was required to keep the reaction vessel temperature below 50° C. The solvent was removed under reduced pressure and the remaining amber syrup was treated with acetic anhydride-pyridine, Workup in the normal manner provided a mixture of several products. Apart from acetyl diphenyl-phosphine oxide, the major product was identified by n.m.r. to be 1,2,3, 4,6-penta-0-acetyl-a-D-glucopyranose. Lithium Triphenyltin(33) 72 Following the procedure of Tamborski , 5 g of triphenyltin chloride (Peninsular Research, Gainsville, Fla.) in 15 ml of anhydrous tetra-hydrofuran were added to a well stirred solution containing 1.1 g of * Sample kindly donated by Mr. A. Tracey, Chemistry Department, Simon Fraser University. lithium and 20 ml of anhydrous tetrahydrofuran under an argon atmosphere. The solution was stirred for three hours during which time i t slowly turned a dark green colour. A quantitative yield would give 0.2 g of 3_3 per ml of reagent. However, previous reports indicated that the for-go mation of this reagent takes place in only 50-75% yields. The solu-tion was stored in a well-stoppered flask at 0° C for several days. Lithium Triphenyllead(54) 5 g of triphenyllead chloride (Alfa, Beverly, Mass.) in 10 ml of anhydrous tetrahydrofuran were added to a stirred solution containing 0.5 g of lithium in 20 ml of tetrahydrofuran under an argon atmosphere. An ice bath was used to cool the solution during the addition. The reac-tion mixture was stirred for 5 hours. During this time, the solution turned a dark green colour. Lithium Triphenylsilane(35) 5 g of triphenylchlorosilane (Peninsular Chemicals) in 10 ml of anhydrous tetrahydrofuran were slowly added over a period of 30 minutes to a stirred solution containing 0.8 g of lithium in 40 ml of tetra-hydrofuran. The reaction vessel was kept at 0° C during the addition. Within 30 minutes, the solution became dark green in colour. The reac-tion mixture was stirred for an additional three hours at room temper-ature . The following reactions were basically performed in a similar man-ner. A detailed description w i l l be given for the reaction of 33 with 20a. Triphenyl p. .2:3,5-di-O-methylene-a-D-glucofuranosej 6-C-stannane(36) 104 20 g of 20a in 10 ml of anhydrous tetrahydrofuran were slowly added, using a syringe, to a flask containing 25 ml of 33_ under an argon atmosphere. A slight heat of reaction was detected'. After two days at room temperature the solution was poured into ten times the volume of water and neutralized with ammonium chloride. This mixture was extracted thrice with chloroform. The extracts were dried over sodium sulfate, f i l t e r e d and evaporated under reduced pressure to give a syrup. This syrup was crystallized from benzenerpetroleum ether (30-60) to afford 2.65 g (87%) of 36_; m.p. 109-112°. The product was identified by n.m.r. spectroscopy (see Tables 11 and 12 and Fig. 16C). Triphenyl |l,2,3,4-di-O-isopropyIidene-a-D-galactopyranosej 6-C-stannane (37) 107 1.0 g of 22a in 10 ml of anhydrous tetrahydrofuran and 15 ml of 33 were reacted in the usual manner to afford a syrup. The syrup was separated on a s i l i c a column (1% methanol in chloroform) to give 0.9 g (63%) of an amber syrup. N.m.r. spectroscopy was used to identify this product as 37_. Small amounts 5%) of impurities was s t i l l present. Triphenyl :1,2:3,5-di-0-methylene-a-D-glucofuranose3 6-C-plumbane(39) 104 1.0 g of 20a in 10 ml of dry tetrahydrofuran were, treated with 20 ml of 34. Subsequent workup gave a syrup which was chromatographed using a s i l i c a column (2% methanol in chloroform) to afford 0.2 g (11.3%) of an amber syrup. The syrup was identified by n.m.r. spectroscopy to be 39 (see Tables 11 and 12 and Fig. 16D). Triphenyl ^1,2:3,5-di-Q-methylene-q-D-glucofuranosej 6-C-silane(40) 104 1.0 g of 20a in 10 ml of dry tetrahydrofuran were slowly treated with 15 ml of 35_. The reaction was highly exothermic and the reaction vessel had to be cooled in an ice bath. Normal workup gave an amber syrup. Column chromatography ( s i l i c a gel—1% methanol in chloroform) afforded a poor yield (< 10%) of a product which was identified by n.m.r. spectroscopy to be mainly 40_ (see Tables 11 and 12 and Fig. 16E). Reaction of 34 with methyl 2,3-anhydro-B-L-ribopyranoside(29) 2.0 g of 29_ were treated with 30 ml of 34_ in the usual manner. The mixture was l e f t at room temperature for several days. The solvent was removed under reduced pressure and the resulting solid mass acety-91 lated via the pyridine-acetic anhydride method. Workup in the usual manner afforded an amber syrup. Column chromatography ( s i l i c a ; methanol in chloroform) gave 0.65 g (21%) of a product, 38_; m.p. 90-95°. The *H n.m.r. spectrum of this product showed two acetate methyl groups were now present in addition to a large aromatic proton region. Integration of the aromatic, methoxy, and acetoxy regions of the spec-trum gave the correct ratios for a product arising from the scission of the epoxide ring. The rest of the spectrum appeared relatively f i r s t order, with several resonances containing numerous couplings. A ten-tative assignment of structure was made in the following manner: It was assumed that the hydrogen attached to the carbon bearing the t i n substituent would be shifted to high f i e l d . Also the two H—5 resonances should exhibit a large geminal coupling. These large couplings are evi-dent for the two protons resonating at lowest f i e l d . From this stage, spin decoupling experiments helped to identify the other resonances. The chemical shifts (x) and coupling constants (Hz) for this assignment are: H-l = 5.9; 2 = <0.5; J 2 3 = ? J J ! 4 = l-°'> J i 5 = 2 ' ° H-2 = 5.32; J = <0.5; J_ . = 2.0; J_ c = 1.5 2,3 2> 4 2 » 5 H-3 = 5.9; J 3 4 = <0.5; J 3 5 = <0.5; J3 5 = 2.0 X 2 -H-4 = 5.11; J = 0.5; J =4.5 4,51 4,b2 H-5, =3.98; J =.-10.5 1' 2 H-5 =3.64 This assignment does c o n f l i c t with one's normal expectations f o r the chemical s h i f t s of sugar proton resonances ( i . e . here protons are at lowest f i e l d ) . However, i t i s s i g n i f i c a n t that a large number of long range couplings are present i n , t h i s molecule. A structure which s a t i s -f i e s most of the above data would be a h a l f chair form of t r i p h e n y l Such a structure has a planar "W" arrangement between tt^Jkl^, H^-Hg, H -H , H,-Hr which would r e s u l t i n some of the long range couplings. It should be emphasized that t h i s assignment i s only speculative. Further experimentation w i l l be required to confirm t h i s structure. APPENDIX A An INDOR experiment is just one special case of an extensive range of double resonance experiments. The following discussion represents an attempt to provide a simplistic, but coherent, rationale for the various double resonance experiments which can.be performed on an AX-system, with particular emphasis on the effects of changing the strength ( Y H Q ) Hz of the decoupling f i e l d ( a ^ ) . The energy level diagram for an AX system can be represented as follows: 1 2 1 2 the transitions A , A , X , and X can be labelled according to the spectrum shown. The Hamiltonian describing such a system can be written in the basic form: y£ = w AI z(A) + " X I Z ( X ) + 2 7 r J A X I z ( A ) I z ( X ) [ A ] where the f i r s t and second terms represent the respective interactions of nuclei A and X with the magnetic f i e l d vector I •. Effectively, these two terms are the chemical shift portions of the Hamiltonian =^£. The third term describes the scalar coupling interaction between the two nuclei with the magnetic f i e l d directed in the z direction. In a spin-decoupling experiment, is centered on one or other of the two doublets and a comparatively strong f i e l d is applied (Y^) HZ 2ir in order to "remove" the coupling from the other resonance; "y" is the gyromagnetic ratio of the nuclear .species whose resonance is being i r r a -diated. If irradiation occurs at the X resonance, the direction of pre-cession of the spin vector I Z(X) approaches the xy plane. The spin vec-tor I (A) and I (X) now form a right angle, making the scalar product z z I (A)I (X) equal to zero. Hence, the last term of equation A becomes z z zero, leading to a collapse of the doublet s p l i t t i n g of the A nucleus. Spin decoupling can be visulized as producing degeneracy in the energy level diagram as shown below: Fz A1 1 2 Here A = A and the A resonance is a singlet. Consider now the requirements of a spin-tickling experiment. In this instance, u ^ i s centered on one transition and i t s strength is set within H the range |JAV|> 2 > 1 , . r . _ u A X 1 — j- 7jT-, where 1 S the spin-spin relaxation time. In other words, the strength of the f i e l d is comparable to the line width of the transition being irradiated. For example, in the A X system,, i r -2 1 2 radiation of the X transition results in both A and A s p l i t into doublets with the size of the s p l i t t i n g dependent upon the strength of the applied f i e l d . Here, the vectors I Z ( A ) and I Z ( X ) are not orthogo-nal, resulting in the last term of equation A now being non-zero. Sim-p l i s t i c a l l y , the irradiating f i e l d can be ascribed to causing a break-down of the AF z = -1 selection rule, which is normally operative, such that transitions effectively corresponding to AF^ = 0,2 are also ob-served. A pictoral representation is given below. 1* 2* A .and A represent transitions which correspond to the previously forbidden transitions involving the selection rules F z = 2 and 0 respectively. If the irradiation f i e l d strength is decreased below the level re-quired for the production of a detectable s p l i t t i n g , then the only effect which may be observed w i l l be a change in the absolute intensity of some transitions, via a General Overhauser effect (vide infra). Consider now experiments performed in the INDOR mode of operation, when the strength of the irradiating dield is appropriate for spin-tickling. Irradiation of a connected transition w i l l cause the tran-sition which is being monitored, to s p l i t into a doublet and the recor-der w i l l show a negative-going dip; when the irradiating f i e l d i s re-moved the recorder w i l l return to i t s original height. Thus a l l respon-ses resulting from a "spin-tickling" type of mechanism w i l l be negative-going. If the strength of the irradiating f i e l d is decreased to a level such 2 that ( Y ^ ) ^1^2 ~*' w n e r e i s t n e spin-lattice relaxation time, then . no detectable spl i t t i n g w i l l be induced in any of the connected trans-itions. However, at this f i e l d strength, which approximates to the saturation level for a single transition, significant changes in the populations of the energy levels can occur; that i s , General Overhauser effects w i l l be detected. A diagrammatic representation of the effect ' 2 of irradiating the X transition is given below: The symbols © and 0 indicate, respectively either an increase or a decrease in population as compared with the normal Boltzmann distribu-2 tion. Irradiation of X w i l l cause a slight increase in the population of the upper level, at the expense of that of the lower level. Since 1 2 2 transitions A and A both have an energy level in common with X , they w i l l both experience a change in intensity. The intensity of transition A* w i l l increase from i t s usual value because the decrease of the pop-ulation of i t s upper energy level w i l l "encourage" the upwards transfer 2 of spin population. In contrast to this behaviour, A w i l l show a de-crease in intensity, since the upper energy level i s now more highly populated than usual. Thus for the particular INDOR experiment under discussion here, a positive-going response would be observed at the 1 2 resonance frequency of A , and a negative-going response at A . It follows then, that the type of response detected by an INDOR ' experiment w i l l be c r i t i c a l l y dependent on the power level *-^2^ . Less ~27~ . obvious is the fact that the "intensity" of a response w i l l depend on the rate at which the irradiating f i e l d is scanned. If the scan-rate is too slow, saturation can occur and the intensity of INDOR responses is thereby decreased. Figure 17, A, B and C shows the responses, at sweep rate 100, 250 and 500 seconds respectively, obtained from moni-toring transition 3 in the D-ribose spectrum. A l l other settings assoc-iated with the experiment remained constant. The 250 seconds sweep rate provides optimum int e n s i t y with good resolution while the 500 seconds sweep rate shows the effect of saturation. The effects of t h i s phenom-enon can be p a r t i c u l a r l y marked i n heteronuclear INDOR experiments. <1i 1A FIG. 17. INDOR responses of H resonance obtained by monitoring transition 3 I in the H n.m.r. spectrum of D-ribose (see Fig. 4).(A),(B),(C), recorded at sweep rates of 100, 250 and 500 seconds respectively. APPENDIX B Theory and experiments have demonstrated that scalar coupling con-stants can have either a positive or negative sign. A positive coupling constant has been defined as one which tends to hold two nuclear spins a n t i p a r a l l e l . U s i n g the Dirac-Van Vleck model, i t has been shown 3 that relative signs between couplings in sp systems should depend upon the number of bonds separating the protons.''''''''' Experimentally, relative sign determinations can be carried out by-using either spectral analysis, spin decoupling or spin ti c k l i n g tech-niques. The only requirement being that the spin system examined con-tain at least three mutually coupled nuclei. For 5-hydroxy-1,2,3,4,7-,7-Jiexachloro-bicyclo [2.2.1] hept-2-ene the following spin assignment Table can be constructed for ^ H — c o u p l i n g s . It .assumes that the geminal couplings are negative while the vicinal couplings are positive. X M A Line 1 2 3 4 5 6 7 8 9 10 11 12 X B a B a B a 3 a M a a B 8 B B a a A a 8 a B B B a a From this Table, spin t i c k l i n g line 12 of the A resonance predicts that lines 1 and 2 of the X resonance and lines 6 and 8 of the M reson-ances (i.e. only a transitions) w i l l be affected. The INDOR responses agree with this prediction (see top of Fig. 1). Sufficient information i s present in these INDOR spectra to con-struct the energy level diagram for this AMX system. Depending upon the direction of the response, the monitored and irradiated transitions are either progressive or regressive. Positive-going INDOR responses (i.e. lines 1,6,4 and 11 in Fig. 1) are progressive transitions, while negative-going INDOR responses (i.e. lines 2,8 and 9 in Fig. 1) are regressive transitions. From these responses the following energy level diagram may be con-structed where the x-, y- and z- axes represent the A, M and X transi-tions respectively. REFERENCES 1. F. A. Bovey, in "Nuclear Magnetic Resonance Spectroscopy", p. 115, Academic Press (1969). 2. A. L. Bloom and J. N. Shoolery, Physical Review, 97_, 1261 (1955) and references therein. 3. W. A. Anderson and R. Freeman, J. of Chem. Phys., 37, 85, 2053 (1962). 4. R. Hoffman and S. Forsen, in "Progress in N.M.R. Spectroscopy", Vol. 1, p. 1-204, Eds. J. W. Emsley, J. Feeney and L. H. Sutcliffe, Pergamon Press (1966). 5. E. T. Lippmaa, J. of Structural Chem., 648 (1967). 6. E. B. Baker, J. Chem. Phys., 37_, 911 (1962). 7. E. B. Baker, J. Chem. Phys., 45_, 609 (1966). 8. (a) R. Kaiser, J. Chem. Phys., 39_, 2435 (1963). (b) J. D. Baldeschwieler and K. Kuhlman, J. Am. Chem. Soc, 85, 1010 (1963). (c) V. J. Kowalewski, J. Mol. Spectroscopy, 30_, 531 (1969); 31_, 256.(1969) . (d) D. G. de Kowalewski and R. Loesener, J. Mag. Res., 1_, 209 (1970).• (e) F. A. L. Anet, H. Lee, and J. L. Sudmeier, J. Am. Chem. Soc, 89, 4431 (1967). (f) L. D. Hall, R. Burton and P. R. Steiner, Can. J. Chem., 48, 2679 (1970). 9. (a) R. Freeman, J. Chem. Phys., 40, 3571 (1964). (b) W. J. Horsely and H. Sternlicht, J. Am. Chem. Soc, 90_, 3738 (1968). (c) F. W. Wehrli and W. Simon, Helv. Chim. Acta., 52_, 1749 (1969). (d) R. Freeman and W. A. Anderson, J. Chem. Phys., 39, 806 (1963). 10. F. A. L. Anet and J. L. Sudmeier, J. Mag. Res., 1_, 124 (1969). 11. (a) W. McFarland et. a l . , J. Chem. Soc. (C), 1136 (1969). (b) H. Elser and H. Dreeskamp, Ber., 73, 619 (1969). 12. R. R. Dean and J. C. Green, J. Chem. Soc. (A), 3047 (1968). 13. P. G. Harrison, Organometallic Chem. Rev., 4_ (A), 379 (1969). 14. R. Kosfeld, G. Hagele and W. Kuchen, Angew. Chem. (Int. Ed.), 7_, 814 (1968). 15. (a) W. McFarland, Chem. Comm., 700 (1969). (b) T. H. Brown and P. L. Green, J. Am. Chem.. Soc, 91_, 3378 (1969) . 16. W. McFarland, in "Annual Reviews of N.M.R. Spectroscopy", Vol. 1, p. 135-163, and references therein, Ed. E. F. Mooney, Academic Press (1968). 17. V. J. Kowalewski, in "Progress in N.M.R. Spectroscopy", Vol. 5, pages 1-31,; Editors J.W. Emsley, J. Feeny and L. H. Sutcliffe, Pergamon Press (1969). 18. R.Hoffman and B. Gestbloom, J. Chem. Phys., 40, 3734 (1964). 19. R. Freeman, private communication. 20. R. W. Lenz and J. P. Heeschen, J. Polymer Sci., 5J_, 247 (1961). 21. L. D. Hall and J. F. Manville, in "The Deoxy Sugars", p. 228 (1968) . 22. L. D. Hall,: i n "Adv. in Carb. Chem.", 1£, 51 (1964) and ref. therein. 23. W. W. Binkley, D. Horton and N. S. Bhacca, Carb. Res., 1£, 245 (1969) . 24. R. U. Lemieux and J. D. Stevens, Can. J. Chem., 44, 249 (1966). 25. S. A. Angyal, Angew. Chem. (Int. Ed.), 8_, 157 (1969). 26. L. D. Hall and C. Grant, Can. J. Chem., 48_, 3537 (1970). 27. L. D,. Hall and P. R. Steiner, to be published. 28. E. F. Mooney and P. H. Wilson, Annual Reviews of N.M.R. Spectroscopy, Ed. E. F. Mooney, 2_, 153 (1969). 29. E. G. Paul and D. M. Grant, J. Am. Chem. Soc, 86, 2977 (1964). 30. A. M. White and G. A. Olah, J. Am. Chem. Soc, 91_, 2943 (1969). 31. (a) L. D. Hall and L. F. Johnson, Chem. Comm., 509 (1969). (b) D. E. Dorman and J. D. Roberts, J. Am. Chem. Soc, 92, 1355 (1970). (c) A. S. Perlin, B. Casu, and H. J. Koch, Can. J. Chem., 48_, 2596 (1970). 32. P. J. Banney, D. C. McWilliams and P. R. Wells, J. Magn. Res., 2_, 235 (1970). 33. R. Kosfeld, G. Hagele and W. Kuchen, Angew. Chem. (Int. Ed.), 7_> 814 (1968). 34. H. Elser and H. Dreeskamp, Berichte der Bunsengesellschaft, 73, 619 (1969). 35. F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, Interscience, (1962). 36. L. D. Hall, C. A. McDowell, R. McLean, and P. R. Steiner, to be published. ;.-37. A. W. Douglas, private communication •38. R. Burton and L. D. Hall, Can. J. Chem., 4.5, 59 (1970). 39. W. J. Horsely and H. Sternlicht, J. Am. Chem. Soc, 90_, 3738 (1968). 40. H. 0. House, "Modern Synthetic Chemistry, Benjamin (1965). 41. H. C. Brown and B. C. Subba Rao, J. Am. Chem. Soc, 78_, 5694 (1956). 42. G. W i t t i g , Angew. Chem., 68_, 505 (1956). 43. E. J . Corey and G. H. Posner, J . Am. Chem. S o c , 89, 3911 (1967); i b i d , 90, 5615 (1968). \ . 44. R. E.. Reeves, Adv. In Carb. Chem., 6, 107 (1951). 45. A. Rosenthal, Adv. i n Carb. Chem., 23_, 59 (1968). \ • , 1 .• 46. G. R. I n g l i s , J . P. Schwarz and L. McLaren, J . Chem. S o c , 1014 (1962). 47. M. M. Kreeyvoy and J . F. Schaeffer J . Organometal. Chem., 6_, 589 (1966). 48. L. D. H a l l and P. R. S t e i n e r , unpublished o b s e r v a t i o n s . 49. A. J . K i r b y and S. G. Warren, "Organic Chemistry o f Phosphorous", E l s e v i e r (1967). 50. J . D. Watson i n " Mo l e c u l a r B i o l o g y o f the Gene", p. 48, Benjamin, (1965). 51. F. Maley and H. A. Lardy, J.' Am. Chem. Soc,'78, 1393 (1956). 52. W. E. Harvey, J . J . M i c h a l s k i and A. R. Todd, J . Chem. S o c , 2271 (.1951). ' 53. M. L. Wolfrcm, J . Am. Chem. S o c , 6£, 23 (1942). 54. B. S. G r i f f i n and A. Burger, J . Am. Chem. Soc., 78, 2336 (1956). 55. R. L. W h i s t l e r and C. Wang, J . Org. Chem., 33_, 4445 (1968). 56. L. Evelyn, L. D. H a l l , P. R. S t e i n e r and D. H. Stokes, Chem. Comm., 576 (1969). ' • ' . '57. L. Evelyn, L. D. H a l l , P. R. S t e i n e r and D. H. Stokes, t o be . pub l i s h e d . 58. G. H. Jones, E. K. Hamamura and J . G. M o f f a t t , Tet. L e t t e r s , 5731 (1968); G. H. Jones and J . G. M o f f a t t , J . Am. Chem. S o c , 9£, 5337 (1968); G . H. Jones, H. P. A l b r e c h t , N . P. Damodaran and J . G. M o f f a t t , i b i d , 92-, 5510, 5511 (1970). 59. R. L. Whistler, C.Wang and S. Inokawa, J. Org. Chem., 33, 2495 (1968). 60. K. Kumamoto et. a l . , Bull. Soc. Japan, 42_, 3245 (1969). 61. M. S. Kharasch and 0. Reinmuth, "Grignard Reaction of Nonmetallic Substances", Prentice-Hall (1954). 62. I. M. Downie and G. Morris, J. Chem. Soc, 5771 (1965); G. Kosolapov and R. F. Struck, ibid, 3950, (1959); H. R. Hays, J. Org. Chem., 33, 3690 (1968). 63. W. A. Bonner and C. D. Hurd, J. Am. Chem. Soc, 67_, 1972 (1945). 64. H. R. Hays, J. Org. Chem., 33, 4201 (1968). 65. J. S. Burton, W. G. Overend and N. R. Williams, J. Chem. Soc, 3433 (1965) and references therein. 66. F. A. Cotton and G. Wilkinson, in "Advanced Inorganic Chemistry", Interscience(1962). 67. A. Feast, W. G. Overend and N.. R. Williams, J. Chem. Soc, 7378 (1965). 68. M. Sharma and R. K. Brown, Can. J. Chem., 44_, 2825 (1966). 69. K. Issleib and A. Tzachach, Ber., 9_2_, 1118 (1959). 70. H. Gilman and D. Wittenberg, J. Org. Chem., 23_, 1004 (1958). 71. A. M. Aguiar, J. Beisler and A. M i l l s , J. Org. Chem., 27, 1001 (1962). 72. C. Tamborski, et. a l . , J. Org. Chem., 27, 619 (1962). 73. K. Isslerb and H. 0. Frohlich, Z. Naturaforsch., 14b, 310 (1959). 74. D. Wittenberg and H. Gillman, Quart. Rev., 13_, 116 (1959). 75. S. 0. Grim, R. P. Molenda and R. L. Keiter, Chem. and Ind., 1378 (1970). 76u. J. Ward, private: communications. , % , 77. D.. H.. Ball and F. W. Parrish, in "Adv. in Carbohydrate Chem.","24, 139 (1969) and references therein. 78. D. E. Corbridge, in "Topics in Phosphorus Chemistry", Vol. 6, p. 258-265, Ed. M. Grayson and. E. J. G r i f f i t h , Interscience (1969). 79. (a) N. F. Taylor and P. W. Kent, J. Chem. Soc, 872 (1958). (b.) J. H. Westwood, et.. a l . , J.. Org. Chem., 32_, 1643 (1967). (c)..R. S. Tipson, in "Adv. in Carbohydrate Chem., £, 107 (1953). 80. J. Cox and L. N. Owen, J. Chem. Soc. (C), 1121 (1967). 81. W. A. Szanek and J. K. N. Jones, Can. J. Chem., 45_, 4223 (1965). 82. A. C. Richardson, Carb. Res., 1_0_, 395 (1969) and references therein. 83. . T. D.. Inch, in, "Annual Review of N.M.R. Spectroscopy", Vol. 2, p. 35-81, and references therein, Ed. E. F. Mooney, Academic Press (1969). 84. C. Cone and'L. Hough, Carb. Res., 1_, 1 (1965). 85. R. M. Hoskinson, J.. Chem. Soc, 2991 (1962). 86.. P. R. Steiner, M. Sc. Thesis, U.B.C. 1969; L. D. Hall, P. R. Steiner and C. Pedersen, Can. J. Chem.,48, 1155 (1970). 87. N. A. Hughes and P. R. Speakman, J. Chem. Soc, 2236 (1965). 88. L. D. Hall and P. R. Steiner, Can. J.Chem., 48, 451 (1970). 89. F. H. Newth, Quart. Reviews, 13_, 30 (1959). 90. J. A. M i l l s , - i n "Adv. in Carbohydrate Chem., 10_, 51 (1955) and references therein. 91. M. L. Wolfrom and A. Thompson, in "Methods in Carbohydrate Chem.", Vol. II, p. 211, Academic Press (1963). 92. P. Karrer and A. Boettcher, Helv. Chim. Acta, 36, 570 (1953). 93. B. Helferich and A. Gnuchel, Ber., 71_, 712 (1938). 94. R. U. Lemieux and G. Huber, Can. J.Chem., 31_, 1040 v(1953). 95. J. B. Lee and T. J. Nolan, Tetrahedron, 23, 2789 (1967). 96. C. Benezra and G. Ourisson, Bull. Soc. Chim. Fr., 1825 (1966); 2789 (1967). 97. A. A. Bothner-By and R. H. Cox, J. Phys. Chem., 73_, 1830 (1969) and references therein. • 98. R. J. Abraham and J. A. Pople, Mol. Phys.,3, 609 (1960). 99. H. Gilman and G. D. Lichtenwalter, J. Am. Chem. Soc, 80_, 608 (1958); H. Gilman, M. V. Georges and D. Peterson, ibid, 82_, 403 (I960). 100. J. J. Eisch in, "The Chemistry of Organometallic Compounds", Chap. 4, Macmillan (1967). 101. H. Gilman, 0. L. Marks, W. J. Tredka and J. W. Diehl, J. Org. Chem., 27_, 1260 (1962) . 102. J. M. Mailan and R. L. Bebb, Chem. Reviews, 693 (1969). 103. M. Barczai-Martos and F. Korosy, Nature, 165, 369 (1950). 104. 0. T. Schmidt et. a l . , Chem. Berichte., 86_, 741 (1953). 105. F. G. Mann and I. T. Millar, J. Chem. Soc, 3039 (1952). 106. F. D. Cramer, in "Methods in Carbohydrate Chem.", Vol. II, p.244, Academic Press (1963). 107. K. Freundenberg and R. M. Nixon, Ber., 5j5, 2119 (1923). 108. G.J. Robertson and C. F. G r i f f i t h , J. Chem. Soc, 1193 (1935); D. A. Prinz., Helv. Chim. Acta., 24_, 1 (1946). 109. T. Golab and T. Reichstein, Helv. Chim, Acta., 44, 615 (1961). 110. A. D. Cohen, R. Freeman, K. A. McLauchlin and D. H. Wiffen, Mol. Phys., 7, 45 (1964). 111. L. D. Hall and J. F. Manville, Carbohydrate Res., 8_, 295 (1968). 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items