The Crystal Structure of Feist's Acid Citation Petersen, Donald Ralph (1955) The Crystal Structure of Feist's Acid. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/SXC6-W012. https://resolver.caltech.edu/CaltechETD:etd-01202004-095607 Abstract NOTE: Text or symbols not renderable in plain ASCII are indicated by [...]. Abstract is included in .pdf document. The crystal structure of Feist's acid, 1-methylenecyclopropane-trans-2,3-dicarboxylic acid, has been examined. The substance crystallizes in the triclinic system with space group [...]. Two enantiomorphous molecules of formula [...] are in the unit cell; the direct-cell constants are [...] and [...]. There is good cleavage at the face (010), and this is also the twinning plane. The intensity data from equi-inclination Weissenberg photographs about each crystal axis were utilized in selecting and then refining a trial structure. Punched-card computing techniques were used to carry out least-squares refinements of atomic positions; two Fourier summations aided in final refinement. All refining work was based on full three-dimensional data. Hydrogen bonds bind individual molecules into chains which extend in the [...] direction through the crystal. The methylene groups extended in the [...] direction form cogs on the chains, and two chains related by centers of symmetry are intermeshed closely. Item Type: Thesis (Dissertation (Ph.D.)) Subject Keywords: (Chemistry and Physics) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Minor Option: Physics Thesis Availability: Public (worldwide access) Research Advisor(s): Sturdivant, James Holmes Thesis Committee: Unknown, Unknown Defense Date: 1 January 1955 Record Number: CaltechETD:etd-01202004-095607 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-01202004-095607 DOI: 10.7907/SXC6-W012 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 246 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 22 Jan 2004 Last Modified: 06 Jul 2023 19:51 Thesis Files Preview PDF (Petersen_dr_1955.pdf) - Final Version See Usage Policy. 3MB Repository Staff Only: item control page

THE CRYSTAL STRUCTURE OF FEIST'S ACID Thesis by Donald R. Petersen In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy California Institute of Technology Pasadena, California 1955 ACKNOWLEDGEMENTS The author is very grateful to Professor J. H. Sturdivant for generous and patient assistance throughout this study and for helpful criticism during the preparation of the thesis. Thanks are due to Professor J. C. Speakman for making known to the author unpublished experimental results and to Professor M. G. Ettlinger for supplying samples of the material studied. Fellowship grants from the Allied Chemical and Dye Corporation and from E, I. du Pont de Nemours and Company are acknowledged with thanks. - ABSTRACT The crystal structure of Feist's acid, l-methylene- cyclopropane-transe2,3-dicarboxylic acid, has been examined, The substance crystallizes in the triclinic system with space group Pl. Two enantiomorphous molecules of formula CgH,O, are in the unit cell; the direct-cell constants are a=4.84 R, b= 9.635 R, c= 7,60 &, “=91.7°,6= 102.1°, andy =110.1°. There is good cleavage at the face (010), and this is also the twinning plane. The intensity data from equi-inclination Weissenberg photographs about each crystal axis were utilized in selecting and then refining a trial structure. Punched-card computing techniques were used to carry out least-squares refinements of atomic positions; two Fourier summations - aided in final refinement. All refining work was based on full three-dimensional data. Hydrogen bonds bind individual molecules into chains which extend in the ¢ direction through the crystal. The metnylene groups extended in the QR direction form cogs on the chains, and two chains related by centers of symmetry are intermeshed closely. TABLE OF CONTENTS INTRODUCTION CHEMICAL BACKGROUND EXPERIMENTAL PART lL. Crystal morphology 2. Accurate cell constants 5S. Intensity data 4, Space group THE TRIAL STRUCTURE REFINEMENT OF THE TRIAL STRUCTURE 1. Application of least-squares methods 2. Fourier summations 3. Correction for anisotropic scattering 4. Correction for scattering by hydrogens RESULTS AND DISCUSSION 1. Configuration of the molecule Le Intermolecular contacts 5. Significance of results APPENDIX A Trial structure proposed by Speakman APPENDIX B Values of the observed and the calculated structure factors Page 50 55 ov 59 43 47 49 52 REFERENCES PROPOSITIONS Page 94 98 INTRODUCTION In 1893, Feist? foune a previously unknown dibasic acid CgH.O,.- The method of preparation indicated it was a derivative of eyclopropene?>* (Figure la,b). Later $ proposed a structure with a central hydrogen investigators atom (Figure lc), which symbolized the unexpected stability 4 of the acid. The most recent suggestion (Figure 1d) was based on a convincing chemical proof. Each of these Figure 1 Structures ascribed to Feist's acid oo Zi CooH _ CCOOH _SCOOH _ _- GHCOOH CHs-Co CHs-CH || CHs-CH| CHy=Co CHCOOH “CCOOH CCOOH CHCOOH Asymmetric Symmetric Semi-aromatic Methylene a b Cc a structures has been challenged, however, and the evidence given by the principal investigators seems in part contra~ dictory. A full crystal structure determination was called for, not only to solve the puzzle of the configuration of Feist's acid, but to add to the information on small-ring compounds. Few complete structure determinations of compounds con- taining three-membered rings have been carried out. The z effect of substituent groups on the dimensions of small rings is not well known experimentally, and the bond lengths and bond angles in such small-ring compounds are of theoretical interest. An investigation based on analyses of the complete diffraction data from crystalline Feist's acid was begun. The methods used in this investigation and the results obtained are described in the following sections. CHEMICAL BACKGROUND The acid OgH.0, was first prepared by Feist™ when he treated 5-bromo~—4, 6-dimethyl-5-—carbethoxy-2-pyrone with strong potassium hydroxide solution. He suggested that this pyrone retained the properties of a substituted acetoacetic ester as well as those of a §-lactone and so proposed the steps below for the reaction with the base, CH CH | CH 3 3 3 \ WN N tou” “yer moog” Nesr Boog pr OH OK 0 OK oc 064 0 i WAN I/\ll Boog * = col Solox H . Hse c=0 0 OK The resulting form of the molecule is the asymmetric, shown in Figure la; an alternate set of steps suggested by Feist leads to the symmetric form, shown in Figure lb. Other preparations have been described; the diethyl ester of eed -dibromo-g-methylglutaric acid, wnen treated with base, forms the salt of Feist's acid’, which is then easily changed 4 to the acid itself. It was also founa® that the isomeric saturated acids analogous to Feist's acid, when brominated and then treated with sodium amalgam, each gave the acid melting at 200° described by Feist. Feist! isolated two substances of formula CgHeO4. Both forms were said to be unsaturated dibasic acids, but several striking differences were observed. He found that the solubilities of the two acids in ether were quite different, that one of the acids sublimed while the other did not, that the crystal forms of the calcium salts of the acids were entirely different, and that one of the acids melted at 200° but the other at 189°, No number of recrystallizations from water served to raise the latter melting point to 200°, Feist implied that the 189° form and the 200° form corre- sponded to the symmetric and the asymmetric structures. He later reported’, however, that both of these forms could be resolved into optical isomers and eliminated the symmetric form of the molecule as a possible structure. Goss, Ingold, 5,8 then showed the 189° form to be an impure and Thorpe sample of the 200° form. The novel semi-aromatic structure proposed for Feist's acid was the outgrowth of early investigations of glutaconie 9 acid . The unusual properties of glutaconic acid and its derivatives, now explained on the basis of tautomerism and 10,11 isomerism » Led to the use of unorthodox formulas to represent the structure of such compounds. To Feist's acid were erroneously ascribed similar properties; these were indicated by writing the structural formula as shown in Figure lc. The hydrogen atom in the center of the ring apparently was meant to represent not only tautomeric . behavior but also the unusual stability and relative inert- ness characteristic of aromatic systems. Feist’, and later Kon and Nanjit®, showed that no such representation was necessary to account for the reactions of Feist's acid and that the acid had no semi-aromatic character as described above. The resolution of Feist's acid into two optical isomers’»8 also eliminated the semi-sromatic structure as @ possibility. In an addendum to one of their papers, Goss, Ingold, and Thorpe” discussed the possibility that the dietnyl ester of Feist's acid, and presumably the acid itself, might have the methylene structure, as in Figure 1d. They treated the diethyl ester with ozone; the decomposition of the resulting ozonide gave no formaldehyde, a product which they expected if tne acid has the methylene structure. Actually, they found diethyl acetylglyoxaloacetate and deduced the asymmetric structure to be the correct one, 0 O3_ Oo fl ZG O00R 0 | -CCooR || CCcooR H=-C\ | 3 CHz- Cx | 1 CHg-C. | CHCOOR CHCOOR: CHCOOR 6 Kon and Nanjir’ performed this experiment and found as | hydrolysis products acetic acid, oxalic acid, and diethyl | acetylglyoxaloacetate. Speakman and his colleagues’” carried out the ozonolysis and found no formaldehyde. On the basis of such evidence it is almost without doubt that ozonolysis breaks the three-membered ring of Feist's acid. Since ozone is such a soecific and effective reagent for breaking double bondst4, this was thought to be strong evidence for the asymmetric structure. Recently a very convincing proof of the methylene structure was proposed’. When Feist's acid CeHe0y is brominated, one of the products is a bromo lactone monoacid CgH.O,Br. Reduction of the bromo lactone by sodium amalgam gives a lactone acid CoH.0, quite different from the original Feist's acid. On oxidation this lactone acid gives trans-cyclopropane-1,2,3~tricarboxylic acid; the original Feist's acid is required to be a methylenecyclo-= propane derivative. Corroboration of the methylene structure came from an analysis which showed that Feist's acid con- tains no C-methyl group. The methylene structure is established by the crystal structure determination to be described. The anomalous | results of the ozonolysis experiment might be explained by the conjunction of the three-membered ring and the double bond, although such abnormalities are rare in the addition of ozone to compounds containing double bonds. The results 7 of the ozonolysis of the comparable substance methylene- cyclopropane are inconclusive, since formaldenyde was the only identified product; the yield was 2% of the theoretical?®, A possible reaction path for the abnormal addition of ozone to Feist's acid has been suggested, but no experimental work has been done as yet to examine the . . 16 | suggestion”. EXPERIMENTAL PART 1. Grystal morphology A sample of crystalline Feist's acid, melting at about 200° » Was obtained through the kindness of Professor M. Ge Ettlinger. The crystals are colorless and rather brittle and dissolve easily in water. Recrystallizations were carried out from water and from formic acid; both batches of crystals melted at 198~199°, and powder samples from each gave the same x-ray diffraction pattern. The erystals used in the subsequent x-ray analysis were grown from saturated water solutions allowed to evaporate slowly at room temperature. Individual crystals, usually split from twins, were mounted in the customary way in order to examine the principal crystallographic zones. Neither the optical goniometer nor Laue photographs showed that any symmetry elements were present, other than perhaps a center, which indicates that Feist's acid crystallizes in the triclinic system. The most obvious crystallographic zone axis, snown as a in Figure 2a, was found to have an identity distance of about 4.8 A. This was chosen as one of the unit cell axes, and two others were selected in morphologically convenient directions. The crystals grown from water were found to be almost Qo Ao a Figure & lized singles er ieal twinned ery 10 invariably twinned on the (010) plane, as shown in Figure eb. This plane also is the principal cleavage plane, although minor cleavage is shown more or less parallel to (100), giving a fibrous type of cleavage when the crystal is carefully stressed in an arbitrary direction. The | brittleness of the crystals tends to obscure this fibrous Cleavage. A saturated water solution of Feist's acid, when examined with a polarimeter, showed no significant optical rotation, for with a tube length of two decimeters, the specific rotation at 20° for sodium D light was found to be less than 0.4°. The solution must be racemic, since the acid can be resolved into two optically active forms, each with large specific rotation’»®, 2. Accurate cell constants In the triclinic system, the expression relating the reciprocal cell parameters to the observable ® is 4”sin®p= nat 4 n-peR4 9 o4? 4 onkasbtcosyt + 2hhatc*cosa¥+2kQb* c#cosod Here o is the Bragg angle of reflection for x-rays; A is the wavelength of the x-radiation; h, k, and & are the reflection indices; and the starred symbols represent the six reciprocal-cell parameters. The six reciprocal-cell parameters were measured approximately from Weissenberg 11 photographs taken about the three principal crystal axes. The accuracy of these measurements was estimated at one to two percent. From these approximate reciprocal-cell parameters, a value of sin” © was calculated for each | reflection nkaQ. Single-crystal rotation photographs around each principal axis were made in a cylindrical camera of radius 50.0 mm and with very low eccentricity. Copper K radiation of wavelength 1.5418 R was used. Precautions were taken to secure an accurate value of & for each observed reflection, and for each identified reflection a value of sin®2 was then found. Many of the reflections were identified unambiguously by tne comparison of the rotation photographs with corre-~ | sponding Weissenberg photographs, and the observed and calculated values for sin’ of these reflections were used to find accurate reciprocal-cell constants by a least-squares method. Tnese reciprocal—cell constants and the direct-cell constants calculated from them are listed in Table 1. Table 1 Reciprocal- and direct-cell constants for Feist's acid at=0.2264 Rt a=4.839 % | p#=0.1115 Rt b=9.630 R c#=0,1584 Rt c=7.599 2 ow = 93°37! x= 91°42! @*= 76°27! @ =102° 6! y= 68°59! ¥ =110° 2! eo 12 Tne volume of the anit cell calculated from these constants is 323.0 g°, the density of Feist's acid, calculated on the basis of two molecules of CgHe0,4 in the unit cell, is 1.461 ¢ em™°, The observed density, found by flotation and by displacement, was 1.472 ¢ em7> at . room temperature, *# 5. Intensity data All photographs used to determine intensities were made on a Weissenberg camera of conventional design with unfiltered copper K radiation. Three films were super- imposed in the camera. The film used was Kodak No-Screen Medical X-Ray Type, and it was developed and fixed in the manner described by the manufacturers. Equi-inclination Weissenberg photographs were made about each of the selected crystal axes. Of the: 1450 non- equivalent reflections calculated to lie within the sphere of reflection of copper K radiation, approximately 13500 were identified. A table listing the sets of films and the number of reflections indexed in each is given here, Relatively low overlapping of reflections among the sets is observed be- cause of the low symmetry of the space group of Feist's acid. ** The similarity of the observed and the calculated density suggests that the eyystars contain no water of hydration, as stated elsewhere Tne function of electron distri- bution to be described gave no maxima otner than those associated with the molecule CeHg04- ee 13 Table 2 Tne number of reflections indexed on each set of equi-inclination photographs Zone Number zone Number Ok 2 146 n4Q L131 LkQ 295 atay'd 125 ekg 262 hko 97 hog 85 hkl 174 hig 139 hk2 167 hea 147 hk3 163 h3h 132 nk4 141 A film factor was determined experimentally for x-ray film of the type used; the value found was 3.72t0.10. This agreed well with the value previously used in these Laboratories-®, Intensity strips based on a logarithmic relation were preparea’®, again with three superimposed films. The films were exposed to the reflection O12 from a crystal of Feist's acid for lengths of time proportional to 3.72", where x ranged from 2.0 to 5.0 in increments of 0.2 unit. On the second film each spot was reduced in intensity by a factor of 5.72, so that x for the series on this film ranged from 1.0 to 4.0. The third film showed x ranging from 0.0 to 3.0. The overlapping ranges gave spots on the film corresponding to relative intensities from unity (x=0.0) to over 700 (x=5.0). The principal value of the logarithmic scale based on the film factor is the integral change in value of a given intensity when observed on a different film in the set. The x-numbers are easily changed into relative intensities for a set of films when all visual estimating is 14 done. The reflections of each set of films were used to find visually estimated intensities. The intensity I(hk2) so determined for reflection hk& from the 28 set of films is related to the square of the structure factor |F, (aiea)| © , 19 as shown 3 2 4 2 gine 9 gin 3 PF (nkg)| “= k,I(hkg)*(——=s->) (1- anges ) | ° | J 1+cos 29 sin’ © Here kK, is the scale factor for the 32h set of films, ) is the Bragg angle, Bnd is the equi-inclination angle. The multiplier terms other than k, are the reciprocal Lorentz sined 1+cos“20 calculated for each value of sin® between O and 1 at inter.~- and polarization factors. The factor ( ) was vals of 0.002 unit, and the points were plotted. Similarly 2 sinke 2 (1-7 )* was calculated as a function of sin® for each sin of the eleven non-zero values of sin, and eleven graphs were made. Values of sin were found for each value hkQ, and the modifications to be made for each intensity were read from the proper graphs. Multiplication then gave for each intensity I(hk) the reduced intensity I,(hkQ); this is directly proportional to |F, (ni2)| *. The proportionality constants kj, Koyeeekyseee relating | nI* and I, differ for each set of films and are unknown. The best set of ratios of the type ky /ky was selected by the examination of the reduced intensities of corresponding 15 reflections from various sets of films. Each reduced intensity I, form the jee film was then multiplied by the ratio kj/k,. In this way all values of I,, were related to the corresponding values of FoI * by the same propor- tionality constant. This scale factor Ky, or simply K, was determined by 20 Wilson's method”~. If for any atom the scattering amplitude is f=f, exp(-Bsin°® /» *) where fy is the scattering for the atom at rest and the exponential factor is due to the tnermal vibration of the atom, here assumed isotropic, then it is approximately true that in any small range of sin® KI, =) £,* exp (-2B sin'® Jz 2) where the bars indicate the average is to be taken over the range; the summation is made over the unit cell. A further approximation gives Cena” in(I,/) 5%) = -1nk - 2B(sin*® /,*) For Feist's acid, fourteen overlapping groups each of one hundred reflections gave these average values. A plot of in(I,/ dL2,*) versus sin’ /r* is shown in Figure 3. The slope of the best straight line through the points and its intercept on the ordinate axis gave values of 4.63 and 3.00 pas ON fag ie 09 3 ty oS Cr 0.1 0.2 0.3 0.4 | ine sin® © ! X Determination of scale and temperature > factors by the method of Wilson 17 for B and K, respectively. The scale factor K was then used to find the set of values EAs the list of Bo values is given in Appendix B. Since these structure factors were determined indirectly from observed data, they are identi- ‘fied as observed structure factors Foe 4, Space group Only two space groups are allowed in the triclinic system, Pl and Pl. The former has no symmetry elements whatever, while the latter possesses a center. The. external form of the observed crystals of Feist's acid. was not useful in determining the presence or absence of a center of symmetry, for none of the crystals examined had fully developed faces; most were ill-formed or twinned. ‘The phenomenon of piezo-electricity is sometimes useful in determining the absence of a center. Large erystals of Feist's acid, examined in a modified Giebe- Scheibe*= apparatus*#*, showed no piezo-electric effect, however, and so this experiment was inconclusive. The distribution of reflected intensities is dependent on the presence or absence of a center of symmetry“. The fraction N(z) of reflections whose reduced intensities are less than or equal to a fraction z of the local average ¥% I am indebted to Dr. W. G. Sly for the use of this instrument. 18 reduced intensity is given by N(z)=1l-exp(-z) for a noncentrosymmetric structure and by N(z) = ert(2/2)® for a centrosymmetric structure. These two functions are usefully different in the range between z=0 and z=1l.. In practice, reduced intensities from each of the taree . principal zones of Feist's acid were separated into groups With a small range of sin®, and local averages Tp were found for each group. The fraction N(z) was evaluated in each group at a set of points 0.1 unit apart and ranging between zero and unity. For each value of za value of Nz) was found by averaging over the groups, and these values are listed in Table 3 for the three principal zones; the average values of N(z) are plotted in Figure 4. Table 3 Values of N(z) for the three principal zones of Feist's acid Ok 6287 .545 .447 .515 .557 .595 .623 .654 .668 .702 nog -171 .288 .417 .597 .529 .570 .633 .668 .686 .696 hko .287 .365 .416 .486 .551 .569 .577 .598 .625 .675 Average .248 .336 .427 .499 .546 .578 .611 .640 .659 .691 The results given in Table 3 indicate that each of the pro- Jections has a center of symmetry. In the triclinic system this could occur only with the space group Pl. 19 (Z)N 20 THE TRIAL STRUCTURE Three Patterson projections were made from the Ok2, the nO, and the hkO reduced intensity data; of these, the first was most significant and most easily interpreted. It is shown in Figure 5a. From the definition of the projected Patterson function for this zone“ 7 £6 P(vw)= a > |F(OkQ)| “cos2nlkv +ihw) K=~00 28 -c0 the following expression was derived for space group Pl; P' (vw) contains multiplicative and additive constants of no Significance in the location of function mexima. so P'(vw) =) |F(oK0)| “coseniev wot +) cos2nkv ) |F(oxa)| ~cosankw Q=1 Ku-@o co eo 2 ~) singnky ) |F(oKg)| sineniw Qzt kz ~00 The peak at the origin was not removed, nor were the peaks sharpened. All summations were carried out with Beevers-~ Lipson strips. The function P!(vw) was evaluated with w ranging in sixtiethns from zero to one and, because of the center of symmetry, v going from zero to one-half in six- tletns. The values were then plotted, and contours were drawn at arbitrary, equal intervals. The function maxima represent the ends of projected interatomic vectors rooted at the origin. Each corner of nD foal 0.0 o ° ee the parallelepiped shown in Figure 5a is such an origin. The four non-equivalent peaks lettered A through D were interpreted to arise from the vectors connecting the projected scattering centers shown in Figure 5b; the coordinates for tnese are given in Table 4. Table 4 Coordinates of projected scattering centers y Zz 0.85 1.11 0.81 0.44 0.17 -0.11 0.19 0.56 The positions of the projected scattering centers at the points tabulated above are not difficult to explain if it is assumed that the observed Patterson peaks are due solely to oxygen-oxygen interactions. The approximate equivalence of peak heights in the Patterson projection does not contradict this assignment. Thus, to account for all oxygen atoms, each of the four scattering centers must contain two of the eight oxygens in the unit cell, Hydrogen bonding across carboxyl groups is significant in Feist's acid; this is indicated by a comparison of its hign melting point, almost 200°, with the melting points of the dimethyl and the diethyl esters of the acid, 36° ana 30°, respectively’. in many straight-chain dicarboxylic acids, hydrogen bonds bind individual molecules into chains extending through the erystal@*, This carboxyl-carboxyl 25 bonding of molecules into chains was assumed for Feist's acid. The orientation of the carboxyl groups relative to the crystal faces was determined from several pieces of evidence. The hydrogen bonds probably lie varallel to the principal cleavage face (010); the fibrous cleavage more or less parallel with the ¢ axis made it reasonable to say that the hydrogen bonds and the assumed chains were aligned in this direction. Packing considerations implied that each of the projected scattering centers is due to one carboxyl group with the two oxygens superimposed. This means that the plane containing all four oxygens of a given molecule is approximately the cleavage plane (010). a These four oxygen atoms were then located in the unit cell. Wire models of molecules of 1-methylenecyclopropane- trans-2,3-dicarboxylic acid were prepared on a scale of one inch to the Angstrom. Normal carbon-carbon and carbon~ oxygen bond distances and bond angles were used. Shadows of these wire models were projected onto the walls of a model unit cell. The carboxyl groups were arranged so that their shadows fell on the projected scattering centers.. It was found that this required the plane of the cyclopropane ring to be more or less normal to the plane of the carboxyl groups, so that the direction of the methylene double bond was very nearly varallel to the b axis. Within these general limits imposed on the orientation and the configuration of the molecule, several different ot molecular orientations were investigated. The shadow of the wire model was cast on tne (100) plane and on the (010) plane. Atomic positions were noted and then transferred to square grids; Bragge-Lipson charts were used to eliminate a few of the orientations. The strong reflections which were used as tests are listed in Table 5. Table 5 Strong reflections in the Ok and the hOX& classes used in testing possible trial structures by Bragg-Lipson charts Class OkQ Class hog olo 6012 100 020 O21 103 O30 O12 104 013 022 202 O32 305 204 The projected atomic positions in the (100) plane were determined quite closely as a result of this procedure. Thus approximate y and z coordinates for the ten principal atoms in the asymmetric unit were found. The x coordinates could not be found without ambiguity, since it was not possible to decide the tilt of the plane of the three- membered ring relative to the (100) plane; the chosen molecular model, however, allowed only two sets of x coordinates corresponding to the established y and z coordinates. One of these sets was chosen at random; the 20 atomic coordinates for this trial structure are given in in Table 6. They were rounded off to the nearest 0.05 of a cell edge to indicate the degree of confidence placed in them. Table 6 Atomic coordinates for the first trial structure Atom xX y Zz Cy 0.50 0.50 0.25 Co 0.50 0.40 0.25 | Cz 0.50 0.25 0.15 Cy 0.75 0.25 0.40 C5 0.50 0.25 0.00 | Ce 0.50 0.25 0.50 or 0.25 0.10 0.00 Op. 0.75 0.25 0.60 Os 0.75 0.25 0.00 O4 0.25 0.25 0.50 | This trial structure compared favorably with that inde- pendently proposed by Speakman and his colleagues@°»"5 ©", Although this work is described in Appendix A, the atomic coordinates found by Speakman are listed in Table 7 so that comparisons may be made. The transformation from his coordinate system to the one used in this paper has been made. 26 Table 7 Atomic coordinates for the structure proposed by Speakman and coworkers Atom x 0.42 0.48 0.46 0.64 0.49 0.635 0.29 0.84 0.75 0.56 y 0.48 0.354 0.220 0.25 0.17 0.22 0.09 0225 0.22 0.14 Zz 0225 0.28 0.16 0.32 0.99 0.49 0.85 0.64 0.96 0.51 Speakman's convention c(4) C(3) (5) 0(2) 0(6) o(1) 0(4) o(1) 0(3) 0(2) 27 REFINEMENT OF THE TRIAL STRUCTURE lL. Application of least-squares methods In the structure refinements which will be described, the three-dimensional least-squares refining technique applied to atomic coordinates is the most significant. This method is now often used in the final stages of the refine- ment of a structure, since it has several important advantages over successive Fourier syntheses. The least squares procedure does not require the plotting of peak contours, and the oossibility of weighting the data used makes the method very satisfying, both theoretically and practically”*, In applying the least-squares process to the initial stages of refinement, as well as to the final stages, a measure of its value as a means of refinement is shown. For the space group Pl, this expression may be written for each reflection nkQ: | | N72 wtak = s2(F-F) = oD [Si ani toe 44 85 Az +§: ] where N/72 Fy=2) £5 cos2rr (hx,+ Ky y+ &z;) j=t and KV 525 are coordinates of 28 atom expressed in fractions of the cell edges a,.b, and c, respectively. The total number of atoms in the unit cell is N, and in the re 28 second equation tr, is the scattering factor for the 328 atom, with temperature effect included. The weignt given each equation is we, The term &, consists of higher order terms and is considered to be vanishingly small. The sign found for F, in the second equation is used for both Fy and F, in the first equation. For each reflection hk, the first equation shows a sum of 35N/2 terms equal to a residual AF, The best set of S3N/2 unknowns of the sort Ox; is defined as the set minimizing the sum Dwar, where this sum is made over all reflections. In the case of Feist's acid there are thirty unknowns. The equations of the first type above are reduced to thirty normal equations, the first of which is oF = yA F) (© Doak $3,= LolSh) ant LoBk SE )ay, + DuGE RE as, plus 27 other terms for which i#1. In general, any cross terms such as 2 os (S535) were left out of the calculation. Terms such as d 0 (SZ 2 3a) applying to the same atom were found to be relatively important and were retained. In this way the 30 x 30 determinant required to solve the 50 equations in 30 unknowns was separated into ten 35 x 3 determinants. These sums of 29 products required for the solution of the normal equations were made by using straightforward techniaues on IBM computing equipment. Four least-squares refinements of the atomic coor- dinates were made. Each of these refinements took about thirty hours and required about ten thousand IBM cards. The essential material concerning these refinements is given in Table 8, Table 8 Results of the four least-squares refinements of atomic coordinates Refinement Reflections Average Number Used Coordinate R y wA Fe Change 1 81 0.0106 0.387 3529 2 262 0.0043 0.514 2508 3 1000 0.0081 0.269 1277 4 1000 0.0011 0.265 1229 In each refinement, those reflections with the largest values of Fol were used. The factor Ris the customary reliability factor calculated from those reflections with [FoI larger than one, the average minimum observable. The last column is the sum of weighted values of AF for all reflections. The final atomic position parameters are listed in Table 9. 50 Table 9 Final atomic coordinates Atom x y Zz C, 0.4525 0.4698 0.2314 Co 0.5008 0.3432 0.2357 Cz 0.4044 0.1868 0.1522 C4 0.6705 0.2500 0.5206 C5 0.4816 0.1710 0.9758 Cg 0.6150 0.2043 0.4981 05 0.2722 0.0936 0.8461 On 0.8442 0.2448 0.6523 0, 0.7488 0.2364 0.9600 O, 0.3491 0.1330 0.5134 2e Fourier summations The structure factors calculated from the final atomic coordinates were used, with the corresponding observed structure factors, to calculate a Fourier Fy synthesis and a AF synthesis, both in three dimensions. The full number of observed reflections was used to furnish data. As an example of the sums involved for space group Pl, P<*82>= FAX FChoe)C, + EFlokeyC, + a Flood) Cy + ee [ FChko) +FChKe)]C,¢, + x x [-F(hke) +F(hRoyS, S, + LE LECheg) + echo] CC, +E EL-F Chog) + FchoR JS, Sp + 2 x0 F Coke) + Feokgs] CG +O BI -F cok AY +ECOKQIS,. OK ol FLED Leet) + PUKE) + FCHEL) + ER KONC, Cy hn « + OYE cake) ~ FOnRED+ EAE £9 + FR KOI SSG + x 2 3-0-F¢ hk) + FOWKT) — FCHRL) + ECAR DISC So * SS SLE chk) + ECHL + ECARD FRRO TCS Se Ft The expression for AP is analogous. C,=cos2whx, etc. All Fourier summations were accomplished by means of IBM punched card techniques using the M-card system devised by Professor V. Schomaker“®, These summations were made with the intervals 1/30, 1/60, and 1/60 for the edges a, b, and c, respectively. About thirty-five thousand cards were used in each of the two syntheses, and roughly sixty hours of calculation time and thirty hours of plotting time were required for each. The centers of the peaks in the Fourier Bo synthesis go0 which involved the twenty-seven were found by a metho grid points nearest the apparent peak center and whien assumed the scattering density e to be Gaussian near its maxima. Some difference was expected between the atomic positions from the Fourier Fo synthesis and those determined by least-squares refinements because of the difference in weighting in the two methods. Nevertheless, the two sets of coordinates agree to within 0.001 in most cases. Figure 6 shows a representation of the three-dimensional Fourier synthesis. The contours shown are those lying on planes parallel to (100) passing near the peak maxima. *WO.tg85uy Btqna sed ; ; ; UCALOSTS [ ST TSadsquy sug pues wotass efTane tad uorgoeTs T st FROPUCO : ° : ~ f . ° 2 TIF yet $ F = WOgTL ys i935 3.509% §SS40T. SUL *stsauquds Os LIpTInEg Teuctsusutp-seruy euq FO UOTZRQUesaadey Dad 9 SsINSTaZ 35 The AF synthesis resulted in an interesting distri- bution of Ae: ‘This distribution near the two oxygens of one of the hydrogen bonds is shown in Pigure 7, The negative regions indicate a smaller vibration in the ¢ direction than allowed for by the isotropic temperature factor. To a somewhat lesser extent, this decrease in vibration in the ¢ direction is shown by all the atoms of the structure. The positive regions, which indicate a greater vibration than allowed for in a direction approxi- mately parallel to b, are shown only by oxygen atoms. The hydrogen atom of the hydrogen bond appears as a smear of Low density between the positions occupied by the oxygens. All of the hydrogens appear in the distribution of Ap at a smaller electron density than might be expected at this stage of refinement. Se Correction for anisotropic scattering One of the assumptions made in the determination. of the scale and the temperature factors by the method of Wilson is that this temperature factor is a constant and represents equal root mean square thermal vibrations in all directions for all atoms. That the scattering density, and so the root mean square thermal vibration, is dependent on direction is snown in Figures 6 and 7. As mentioned above, this anisotropy is most pronounced in the ¢ direction. The correction for this was made by superimposing upon the Bs eD Figure 7 0.5 "i Th rom nestroms ex: ubie A&A oF is) i ee +Ha= cubi » % D electron per c Orn oer tom: te 2 2 on in the three-dimensional Pouricr A fed © Rey re fet Go Om rio Gi Pea oO, 42> ord @ ri © i a ri Ary, @ fs So @ oO 4 Gr Bay 42 oe ba oO O @ 35 constant isotropic thermal vibration a linear vibration parallel to ¢. This correction reduced to exp(qQ”), where q is a constant and 2 is the index number for the c direction. The constant q was found empirically. Values of |F,| and |F,| were used to find the relation between the Q index and the deviation of > FI from the corresponding > |F,| for each zone of constant f. Table 10 Relation of sum of |F,| and |Fgl for zones of constant 3 Zone LIFS| YIFel LIF | /dIFeI nko 541 591 0.92 1 720 762 0.94 2 759 855 0.89 3 783 825 0.95 4 443 410 1.08 5 447 578 1.18 6 o41 270 1.26 7 153 105 1.46 8 Le1 68 1.678 9 54 27 2200 A graph of LIF LI /LIF I as a function of 2” must be fitted by an exponential function of the type exp(aQ”). By trial, & value of q=-0.01 was chosen, giving the curve shown in Figure 8 A slight change in scale factor was also called for. These changes were made to give the values in Table ll. 80 Q 70 nh 40 50 60 ¥ Fi 30 re —d » f 2.00 1.50 1.00 alr 2 IF, 37 Table 11 Relation of sums of |F,| and |F,| for zones of constant 2 after anisotropy “correction Zone FIFI d IFeol LIF /LIFel hko 540 544 0.99 1 818 779 1.05 ra 863 901 0.96 3 880 902 0.98 4 497 482 1.03 5 502 485 1.04 6 583 388 0.99 7 172 172 1.00 8 136 128 1.06 9 60 60 1.00 The anisotropy correction reduced R and » wAF@ to 0.181 and 1070, respectively. 4. Correction for scattering by hydrogens Approximate positions for the six hydrogen atoms in the asymmetric unit were calculated on the basis of geometrical considerations and previous determinations of structures having carbon-hydrogen bonds. The hydrogen bonds were assumed to be symmetric. These considerations gave the hydrogen positions listed in Table 12. 58 Table 12 Approximate atomic coordinates for the hydrogen atoms Atom x y Zz (Cy) Hy 0.279 0.493 0,154 (C1) Ho 0.582 0.565 0.325 (Cz) Hg 0.180 0.130 0,130 | | | (C4) Hy 0.900 0.250 0.520 | (0,) Hs 0.511 0.114 0,680 | (O05) Hg 0.797 0.241 0.796 The corrections to the calculated structure factors due to scattering by hydrogens were found. The isotropic temperature factor B= 4.63 was assumed for the hydrogen atomse It was not necessary to calculate corrections for those reflections with sin9>C.5, for scattering here is reduced to 0.1 for each hydrogen. The final values of R and > wAF” were found to be 0.172 and 962, respectively. Tne resulting sets of IFO and FP, are listed in Appendix B; the corrections to values of F, due to scattering by hydrogens are listed as F(H). 59 RESULTS AND DISCUSSION 1. Configuration of the molecule All bona lengths and bond angles described below were calculated using the final set of atomic position parameters listed in Table 9 and the direct~cell constants listed in Table 1. For the unit cell defined in Table 1, the square of the bond length is n= 25.415(xy-x4)"+92.751(y;-y3)" +57. 743( 24-25)" ~ 52.089 (x4-x 5) (y4-y;) - 15,418(x4-x 5) (24-25) ~4.355 (yy -y 3) (24-25) where X4Y42Z, are the coordinates of the jth atom. The values found for the bond lengtns are listed in Table 13. The bond angles are also listed in Table 13, Figure 9 gives the identification scheme used in this table. | | | The standard deviations @ of bond lengths were found by the method of Cruikshank®!, The standard deviations of the bond angles were derived from these. It is assumed conventionally that discrepancies of t3@ may be expected. The average standard deviation for bond lengths is near 0.01 2 ana for bona angles near 2°, a 40 Table 13 Bond lengths and bond angles for the Feist's acid molecule C4-Cy 1.517 & 5-01 1.253 3 Co~C 1.497 & C50, 1.261 2 Cp-C4 1.490 & Cg-O5 1.275 & CnC y 1.545 Cg-04 1.263 2 Cx-C5 1.482 & 01+-.0, 2.652 2 C4-Cg 1.480 2 Oo.-.0g 2.625 & 0} -C,-C. 146°35! C,-Co-C, 150°40! Co-C Ce 114°51! | O50 ,-Cg 115° 3! CO 4-Cg 117°44! C4-C,-C., 117°12! CaO o-0 4 62°45! Co-Ca-C, 58°48! Co~C4-C 2 58°27! Ca=C5—05 116°56! C,-C.-05 116°49! C=C 5-0. 120°27! | C4-C =O, 120°18! 01-C,-0. 122°37! Op~C 4-0, 122°53! ' | ; . PTOY 8,98Teg JO eTNosToOuU suy uy se~sue puog pues suy Tt 41 eSIl dl 42 Tne carbon atoms of the cyclopropane molecule form an equilateral triangle with sides of 1.52 go*. The bond lengths between atoms in the tnree-membered ring of Feist's acid suggest that this triangle is not equilateral, even | though the differences from 1.52 R are within the t+3e limit. The general agreement between equivalent parts of. the molecule unrelated by symmetry elements of the space group is remarkably good, and for this reason high confidence is placed in the measured bond lengths and bond angles. The slight shortening of the carbon-carbon bonds in the ring adjacent to the methylene group might be caused by double bond character, but convincing resonance structures are not easily found; however, structures involving separation of Charge may be written. The carbon-carbon bonds from carboxyl groups are considerably shortened from the normal single-bond distance of 1.54 R. The distance 1.48 R represents about 15% double bond character (on the Pauling scale’) for these bonds, This decrease in bond length has been noticed in other carboxylic aciag?*s 99 and is probably not unusual. The carbon-oxygen bond lengths in the carboxyl groups are puzzling. All of the bonds are very nearly the same length, 1.26 A, which is about the length expected if each of the carbon-oxygen bonds had 50% double bond character. It was suggested°° tnat this might be indicetive of a random arrangement of pairs of carboxyl groups in the crystal, the 45 randomness being confined to the flipping of the groups through a half-circle. The bond angles in each of the carboxyl groups suggest this is not the case. An empirical relation between the bond angles and the double bond character of the carbon-oxygen single bond in a carboxyl group indicates about 30% double bond character for these bonds in Feist's acid. This is shown in Figure 10; the plotted points were calculated from a dozen recent structure determinations of carboxylic acids@*,57 and the darkened points were determined for Feist's acid. . The double bond is clearly outside of the ring, for the bond length 1.32 a for the 01-C5 bond is very close to the conventional double bond length 1.33 R. The trans position of the carboxyl groups above and below the plane of the cyclopropane ring (Figure 11) also shows that the methylene structure is the correct one, 2. Intermolecular contacts The contact distances between neighboring molecules are of interest. Only those intermolecular contacts on the order of 4 g or less in length are given. Figure 1l shows the method of identifying molecules, and the lengths of the significant intermolecular contacts are listed in Table 14. The primed molecules are those in the layer directly below the unprimed molecules; thus M' is located at the end of the vector -a rooted at the molecule M. The double primed | \\ 0 2 124 hb O ©O oO 4 - o.oo | | 122 |. \ 4 I} \\o ; 120 } a \ O 4 O ~ | \ I O | | i rN ma Wi, 10 20 30 40% “0 DBC \ . O—H - 128 |- O | / | a J C 126 | O 4 - TOF O O/O : 124°} - O /o O -. 4 122 ¢- O / © 4 / 4 . LA i | 10 20 30 40% DBC Empirical relation between bond angles in the carboxyl sroun and double bond character in the carbon-oxygen single bond. Derkened : Sireles are for Feist's acid, cw) <fi aueTS (OOT) euy ut Sutyor eh LL Sanetyg sueTo (OTO) sug ut Sutyoeg (a) LL sansty 46 Table 14 Distances of intermolecular contacts between the molecule M and its neighbors M to A M to B Cy--C; 4.01 C,--C, 3.60 8 Gj--Cg 3.85 A Cj--C5 378K f Cy--02 3.73 8 Cy--0; 4.05 2% | Cy--0, 3.94 8 Cy--03 5.55 A M to E M to F | | Cg--Cg 3.71 8 Cp--Ce 3.59 R | Cy--0, 3.83 2 Cxm-On 3,92 2 Op--0, 3.50 8 Ox--0, 3.52 2 | M to Mt M to Dt CoO, 3.57 A O1--0n 5.16 B CumeOz 3.40 2 Cp--0, 5.79 2 O,--C, 3.85 2 M_to EB! O,--C, 3.82 & O4--0, 3.43 8 O,--0, 3.54 2 | Oy-“0p 5.26 R 47 molecules are in the layer above the unprimed molecules. Individual molecules of Feist's acid are bound into chains by hydrogen bonding across the carboxyl groups; these Chains extend in the ¢ direction, Each chain is built up of only one enantiomorph and is related to a nearby chain made up of units of the other enantiomorph by a center of symmetry. Tne double bond of the methylene group is extended in the b direction, and the resulting shape of the molecule is such that each chain appears to have teeth or cogs if looked Gown upon in the a direction (Figure 10). The cogs of two chains interlock to give tight packing. It is © interesting to note that each methylene carbon C, is nestled 1 in the center of the ring formed by two carboxyl groups in the opposite chain. The van der Waals contact distances of the oxygen atoms are longer than might be expected; the pronounced anisotropy of the oxygens is probably related to this. Also of interest is the root mean square thermal eennaiand vibration of the atoms; in the ¢ red au direction, (ug“)?=0.172 &, a | and in the plene normal to g, (ug”)*=0.243 2. other directions have values intermediate to these, 5. Significance of the results The dimensions of the three-membered ring, the position of the double bond, and the trans configuration of the carboxyl groups show that the methylene structure is correct, The results of the ozonolysis experiment must have been ee 48 interpreted incorrectly. It is suggested that a thorough investigation of the addition of ozone to Feist's acid and to related compounds?® might throw light on the general problem of the mechanism of the ozonolysis reaction. 49 APPENDIX A Trial structure proposed by Speakman During the course of this investigation of the crystal structure of Feist's acid, it was learned”? that a similar investigation was being carried on by Professor J. C. Speakman of the University at Glasgow. Preliminary Fourier projections were shown by him, and a trial structure was based on these projections. The following direct unit cell was given. Table 15 Direct-cell constants for Feist's acid al=4.03 8 x! =95° pi=7.6 & @'=968" ef=9,15 & 3! = 78° with two molecules of formula CeHe04 in the unit cell ana with space group Pl. Table 16 lists the coordinates of atoms in the trial structure proposed by Speakman. 50 Table 16 Trial structure proposed by Speakman Nomenclature x! y! Zz} Nomenclature of this paper of Speakman Cy ~O.16 0.27 0.02 G(4). Co 0.14 0.27 0.16 C(3) Cs 0.2 0.54 0.30 C(5) Cy 0.89 0.18 0.25 0(2) C. 0.32 0.51 0.35 6 (6) Ce 0.41 0.01 0.28 6(2) or 0.20 0.65 0.41 O(4) | Os 0.59 0.14 0.25 o(1) On 0,53 0.54 0.28 0(3) 04 0.22 -0.01 0.36 0(2) The unit cell described by Speakman is related to that used in this investigation (Table 1) by the vector equations 4 c= bt Similarly the reciprocal cells are related by Ho—ualH#4o ote ge Sn-a +e — 1 bem olt — ' ct=— b!# The atomic coordinates in direct space will transform in the 51 same way reciprocal axes do; the transformation of Speakman's atomic coordinates to the system used in this paper was carried out. It was necessary to add $ to each coordinate value so found and to invert each resulting atomic position through @ center of symmetry before comparable values were obtained. The trial structure proposed by Speakman is given in Table 7 on Page 26 and should be compared with the trial structure found independently (Table 6) and with the final atomic parameters (Table 9). 52 APPENDIX B Values of the observed and tne calculated structure factors The number representing the observed structure factor IFQl for the reflection nk& was found from the intensity of the reflected x-ray beam from a small crystal of Feist's acid as described on Page 12. The calculated structure factor Fo is a function of the positions of the atoms in the unit cell and the scattering capabilities of these atoms. Ideally |F,]| and |F,| should be equal for each given reflection, but the limits of experimental accuracy prevent this; some function of the differences Fo-F, for all reflections is usually used as a measure of the quality of a structure determination. The third column lists values of e function F(H); this funetion is the contribution of the scattering by hydrogen atoms to the calculated structure factor. The terms F, in the second column already contain this contribution. Where no value is listed for F(H), it is negligibly small. No values of F(H) were calculated for reflections with sindyo.5, The value of Fl corresponding to the lowest observed intensity averaged somewhat less than unity. Any reflection with an unobservably small intensity was given one-nalf the minimum observable intensity; the resulting values of FSI are all less tnan unity in Table 17. ers) Table 17 Comparison of observed and calculated structure factors | Fol Fe F(H) 002 1 2.05 2.28 «56 2 6.76 6.29 ~3.58 3 8.16 - 8.14 246 6 13.54 12.55 7 3.89 - 4.70 g 5 Ad 6.07 9 3.59 - 2.74 012 1 4,92 ~ 3.24 054 2 19,25 20.12 - ,24 3 10.80 10.68 - .74 4 090 76 - .16 5 64 - 1.43 16 6 17.71 17.71 7 5251 6.36 8 4,24 334 9 1.24 13 O2Q fe) 47.49 -46 64 - 36 1 23.84 22.38 o14 2 10.79 -10.99 - .10 3 14.20 -16,14 - 52 4 3.87 - 3.28 - 30 5 6.39 6.37 7 5.25 4,28 9 4.37 5.08 O5% 1 3.44 2.26 +08 2 9.92 - §.61 -1.68 3 6.29 A, 30 08 4 9.45 -10.47 1.34 5 5.66 ~ 4.96 6 16.48 -16.79 7 3.06 3.59 8 1.50 ~ 1.63 54 [Fol Fo F(H) OAR fe) 5.30 7.80 058 1 5.20 6.26 - 18 2 5.91 4.65 = 06 4 9.12 - §.81 : 5 6.15 - 4,42 | 6 4.60 - 4.38 | 7 2.11 - 2.77 8 63 - 49 | O52 re) 89 228 - .68 1 4.00 BAS 002 2 9.20 9.18 AA 3 5.37 8.21 4A 1.52 1.15 5 1,94 -~ 1.72 6 3.75 4.96 7 2.72 - 2,02 8 1.07 1.52 CBR 1 63 12 24 2 65 - 1.21 3 411 - 5.69 4 7.26 9.13 5 2.45 2.65 6 265 97 7 205 - .75 O72 @) 3.06 ~ 5.12 2 1.74 - .67 3 2.52 - .28 4 3290 3643 5 - 41 6 - 1.23 089 e) 1.07 5350 1 1.56 - 2.20 2 1.96 - 54 3 205 209 5 1.95 6 77 oO \O r=) © ® a @) e oO e a) f~ e a) Oo ON ATE WNH OM WOON AU EW MH OP” VE OP WNHNHOD}O WFD HO 1.75 3.01 1.40 065 63 1.55 47.24 3450 1.20 4.72 18.57 1.70 11.86 2.68 3.09 13.55 23.91 6.85 19.54 4.80 3.44 6.53 6.68 1.19 ‘x 2.68 - 2,22 ' a s \N bt 13.96 25.22 - 7.68 18.47 1.34 593 6.23 1.56 F(H) 210 - 204 92 1,24 = 78 04 - .08 - 18 a) AW) te] bh > 7 WI OUP OIMH Ox. NAUFUWNDHF OO NAM FUNHOG ON OUP U DH OM CON OVUT Ol PO HO |Pol 13.52 2.28 10.79 17.80 6.01 8.12 1.57 1.38 2.40 18 44 264 10.56 7.74 4.33 2.71 1.52 r OV OV 1 i) Mm e BH PUTT HF e e e NOI OW NOON UW OI 1.79 1.05 Fe 2.48 9.73 18.49 549 - 7.94 1.63 239 2.24 -19.46 10.25 ~ 9.06 3.90 e Cd e ARI AEWY WOON EPH NO e @ i MOE OY BO) bs e e 8 e Vl Te) 11.22 - 6.75 - 5.70 3-33 340 1.31 - 2,19 ofl 1.84 1.62 - 1.28 2.72 F(H) -2.82 014 228 76 1.92 12 1.32 226 46 46 238 242 172 WO or © ON OAUTFEUNDEF OR NRFOY WHHOS UEUNE Or 218 N QUI OH O 1,2° 4.84 66 66 9.44 5073 5.91 4,50 385 3.89 5.04 1.28 3.36 2.09 11.12 1.15 2.25 1,62 1.52 1.44 1.63 57 Fe - 2,00 1.25 027 98 096 26 1.73 °77 2.02 64 feed 1.82 4.14 1.20 1.14 i} a] oe U1 tM fr MN EW OO eeeee © @ @ WONNEFAODN AS OV 2 A PO I ON4O 1D - 60 1.18 2.45 1.55 013 ae 1,01 F(H) -1.40 20 60 ~ (0 hy) ne) WN hy) NAN FUME OO AAU EUW MH Ode Nn —- — 4,45 7.93 2.43 5.86 2.10 265 ie) 3), UW EOWNMH OOP ANEW NMHPOO AUNFUWNMHO nm e UI ~] 4459 065 PH AO OM e ND HW CO e NH EA 58 c 13.58 3-30 6.33 1.62 4.70 5.16 2.28 4.18 9.76 375 F(H) .28 .68 056 032 250 59 [Fol Fo F(H) 27Q 8) 1.28 228 | 1 1.99 2.67 2 266 017 3 2.22 2.84 4 80 226 282 @) 1.75 - 2.24 2 2.09 2.12 3 ° 74 - 1.77 222 0 - 12 308 1 9.40 -10.71 2 1.26 1.29 3 3253 - 3.28 5 1,06 - 1.21 6 4,66 4,61 7 1.53 31g 0 1.76 2.40 1 11.47 -10.26 2 1.37 - 1.15 3 8.38 - 9.32 5 3.03 2.93 6 63 220 Dek 0 1.24 1.10 2 2.91 - 2.60 3 64 1.42 4 3.84 - 3.50 5 1.35 - 1.41 6 2.73 - 2.95 60 lI Fo F(H) B3Q 6) 1.08 264 1 14.05 14.16 2 264 - Ad 3 2.98 2.76 5 262 203 BAR 0 3.29 1,66 1 6.27 6.36 2 264 0,00 3 1.70 2.12 4 62 oA7 5 62 62 35x ie) 063 - 90 3 1.15 - 1.16 36Q 1) 2.11 ~ 1,22 2 1.47 267 3 - 298 37h 1 261 057 2 - 04 AOL 0 61 286 1 11.59 11.47 3 2.02 1.49 4 1.43 056 5 260 098 A1Q fe) 1.44 - 1.04 1 5.48 4.00 2 2.32 2,02 3 4.09 3.89 4 260 - 035 5 «60 - 44 61 [Fo | F, F(H) 429 4 0 1,02 224 | l 1.14 - 10 2 2.26 2.02 3 1.67 - 1.51 A3Q 0 261 68 2 1.04 - .¢6 3 1.78 - .c6 AAg 0 2.38 - 1.88 1 059 04 2 «62 50 45 1 059 - .10 508 1 1.23 - 1.33 3 62 - 37 51g 0 1.32 - 88 1 1.29 1.21 52Q | fe) 61 - .26 1 1.70 - 1.78 o1g 2 24.70 25.20 40 3 33.16 ~35.53 290 4 8.96 -~ 9.53 054 6 265 2.55 8 2.74 - 2,62 9 7 AT - 8.14 (e) i) © COON AUT SU Po 20] Oo o) & Ol LF OONDUEUNDH RH O ON DUI EU DD HX! (e) 1) O CON OUT FO! ND RH 14.66 27.02 15.94 1.88 5.45 9425 1.97 1.96 5.11 21.53 14,11 64 10.91 2.51 4,29 3.25 5.40 5.55 16.55 3.58 9.32 4.31 1.91 2.72 065 2.21 5.69 2.31 3.17 1.85 1.98 2.75 63 62 Cc -14 33 24.27 ~16.58 3.00 - 083 - 1.21 14.56 241 11.27 338 - 3.26 -~ 3.64 4,68 - 5.46 21.71 11.27 3 50 - 2,62 1.17 597 1.43 1.34 1.90 - 3.12 2.25 - 3.01 BAAS F(H) 092 1.98 42 - .66 ~ 42 -1.78 260 - .60 058 224 7 Ads el2 63 | lr Fo F(H) | 062 1 1.64 1.31 036 4 4.56 4.81 6 4.59 - 7.14 7 65 1.27 8 2.23 2.43 O7Q 1 4.08 4,34 2 6.46 6.64 3 2.36 2.235 5 1.89 ~ 2,41 | 6 065 2.38 ost 1 3.64 ~ 3.78 2 265 2.06 3 5-29 7.18 4 053 038 7 265 - 091 09g 1 263 240 2 1.40 260 3 3.89 - 4,81 4 1.09 263 6 - 15 0.10.8 1 «63 - .28 2 1.06 052 oO e fH a) e FW py 42] js oO LO CON] OVUT WI pp 0] a OY) OW ON AU FU fo H 2-} 0 ON OUI Ul 10 0} 1.13 065 63 6.67 3.00 27.05 12.09 9.05 7.36 1.91 1.82 2.97 25.76 16.34 37.18 9.40 13.85 6.36 4.82 1.18 3.55 1.91 5-90 6.75 7.65 4.97 1.38 6.88 «66 64 1.47 218 = 15 - 4,40 209 8.00 7 D4 - 2,24 1.35 ~29.22 33.78 10.85 13.36 - 6.36 ~ 5.58 280 1.08 2.93 6.51 7.27 4.38 072 7.21 F(H) 1.36 ALAA 08 ~1.36 04 288 =2.06 - .24 e80 206 2 O44 1.90 18 266 18 - Ol Lb ODIOHUW Iw HS} aa —_ 1O CON) QVUT Ol 1) HO & I CON OUI EWP he] (0 COON OUT EU DH a 3.48 1.29 1,00 4.68 4,87 4.80 264 093 5292 9.03 9.37 4,10 WN nm ar ar a UI AON AN NAAN « HUT OH » UNAM FORO VON EPONA e 65 Fe 3 34 3201 1.67 ~ 4.38 1 5e71 034 1.53 6.12 7.33 9,12 AAG 1.41 1.26 03 3.68 4.23 1.38 6.15 3285 1.72 3.82 3.19 090 7027 70dd 82 732 077 1.47 1.10 F(H) 038 052 228 220 -1,02 Lad ~ co “I NON LUD Ha} ON OUT EO 1p 18 =) Ke) AU Up Hs! =) e _ Ss fl 19 He! iy?) oO (OO ON] AUT & UO! ND FY do] 3208 3.45 4.51 3.58 2.05 266 64 4.72 1.31 2.41 1.75 4.36 287 1.82 256 330 2.55 065 88 1.25 5420 24.16 5.97 13.79 507 1.46 «63 4.53 a ~ ° OI =) 1.27 4.86 2.34 1.35 1.37 - 1.21 024 016 5.41 6.44 15.16 1.46 - 6.39 1.70 057 5.26 F(H) 58. 84 032 i) et AY) ho WON DIEU PL OONAU FUNDED WOON OUI EUW 1 He] ike) ff 10 ON AUT Ol po 201 13,01 14.17 10.33 8.48 4,15 1.72 18.26 8.74 4.70 9-51 5.00 065 4,05 1.63 1.03 6,22 14.56 1.39 13.17 3.80 4,82 053 065 1.77 3023 9.24 2.75 6.89 3293 265 1.93 265 13.24 15.73 7-79 8.24 - 1,09 - 3.52 vey - 8,14 3.03 - 8.63 3.04 = 203 2.12 F(H) 64 - .06 216 - .06 1.16 - 054 - 222 210 ON OUI Ol 19 7! ON OUT FO! 19 Ol nN Pl NOU Ol pr { AUEWD HS! rm Ke) Ul pH ty Qo. ee e¢@ 0 @ Min Oo ONU1O Oo OR OKRFAUOH NM MUTOH HP eH ° 2.00 093 1.56 266 1,06 1,00 | . QO oo e @© @ @® @ MRUIN He Me OVO © 0910 UI PO NAO FROWN nw Ww “J ° 4.66 2,25 8.11 049 3.10 027 434 1.77 2.49 1.49 283 1.31 2.06 1.96 87 1.89 2.12 ed7 1.16 1.50 F(H) WN (oe) 0 ON OW FU pH! WN om Ui nN WOONAWEUD RA 0 ON NU EU fo el Wi WW ON AU UI pp HP! Ww i CONT OV"UT PU pp He! ol 3.05 8.94 11.72 6.76 17.94 3239 2.06 1.77 6.00 7.02 4.57 5.98 13.7 ALAL 1.53 2.24 10.74 5.63 334 1.28 6.48 3.82 9239 2.46 3.80 7250 064 2.45 1.78 1.31 69 Fe 1.56 1.35 5.89 - 2.21 9.07 075 4.90 - 2.05 4,28 - 1.09 - 1.17 1.67 - 1.15 1.50 032 212. WN WI N QUI FOI HS! AT ONT EO fp Ho] W W (ee) “I Worl UO EW Hol tf © CON AUT Ol pO 0] -_ wr CON OO f Ol PO FH I [Fol 1.41 3299 3.14 262 2.26 1.23 2.80 2.16 1.34 262 61 264 264 3478 262 3.81 3219 TAT 3.06 1.66 284 391 5.88 1.35 4.62 3295 70 Fe Ad 5.00 2.14 366 3.00 .60 65 1.79 053 216 044 029 2,06 315 2048 1.95 2.28 49 SI 206 04 3 03 1.10 3.06 3.48 7.86 2.89 1.0° elf / o49 4.67 589 092 3059 3.13 2.55 78 7.82 F(H) = a> wl NOUEUDE™! oNau EU ND Her i ty > Ul OUT Ul pp pe) > ' fon FUN Hest UI EU po He] wm © ST ONUT Ot pp Hol ig Q. 6 Wen NW ° ° AVION LH IA] ON 90 10 60 °e 2.23 262 3259 1.95 1.40 1.75 4,13 262 1.15 260 061 1.54 62 2.44 1.36 1.17 62 590 1.40 3.05 260 097 097 71 3.84 6.66 1,14 1.29 3.11 003 095 1.14 1.70 258 3.02 1.15 1.12 1.73 1.68 4,58 085 1.31 208 095 220 019 1.721 1.67 1.39 206 1,02 092 o77 095 1.42 - 2.56 212 1.77 043 075 137) jt ONUT Fi no HDI An nN OV Ol po He! Ul Ol Ul 1 Hel my jt 1 wl OONNDVIFOD EH OF OONADAUEUWYDHO © 059 2.52 262 1.74 059 262 1.32 ~60 °61 1.49 ~62 2.00 21.75 6,12 12.14 16.09 1.17 4.82 6.66 6.53 1.97 29.06 21.57 16.09 9.60 2.71 1.24 15.61 3.85 2.42 72 Fe 1,09 1.83 2/0 2.25 1.16 2.38 293 «63 -16 054 003 1.35 054 2.19 O07 17.58 3.58 12.56 15.59 1.20 4,98 7.68 5.61 2.50 2.56 26.48 20.78 17.45 9.91 2.08 oe) 16.91 4.50 2.20. 2.29 F(H) 1.42 3134 “40 202 - 88 1.00 «94 - .92 - 46 Gl i er 1 Ov CONN FUWNMH Om ONAULPU DEH Or 61.61 9.17 15.39 4.53 4.74 9.03 15.74 266 3.10 14.01 5679 6.17 10.17 1.76 4.48 5.54 3207 1.03 21.01 748 13.04 10.41 2.17 5.31 2.98 2.96 19.14 4,12 8.96 4.85 3.72 4.42 3.17 266 3265 73 F(H) 1.02 210 296 - 48 48 -1.80 - 34 66 -16 »26 42 - 78 048 oe] OO} FUDD ORF ON NUIFU DH Ox =) 2 6l ° FUND HOS AUFU MDH OX AUFUNMF OMS NaU [Fol 5.73 2.09 8.13 731 4.93 1.54 1,66 1.57 3.00 4.65 4.35 2.40 1.27 2.17 266 2.28 1,83 5.28 2.07 264 1.81 065 64 66 1.23 64 065 83 064 1.05 266 264 74 Fe 6.84 05 11.09 552 1.15 88 030 3.62 5.53 2,62 - 43 - 1.11 5.52 - 02 216 - 2,72 - 1.52 - 02 46 057 87 - 1.31 ~ 054 054 1.47 219 F(H) nm | { ONAN EVYDH OF ONATELUDH OX NM sxe re) ONIATEFUOUNHF OV” ON NOU FU MrHO lFol 31.40 7.99 13.46 10.77 4.39 4d 2.46 8.34 10.38 17.50 2.91 6.79 4.90 1.77 9.96 3.14 10.38 4.50 7.84 3.51 1.87 323 11.58 8.89 5.18 65 75 Fo 27 58 8.09 12.52 10.37 2.56 - 4,44 1.60 - .07 - 1,40 ~ 2.04 10.83 -~ 9.14 17.95 2.49 4.65 A Ad 1.93 - 4.54 2.70 - 6.86 035 “11.11 1.98 7.28 203 ~- 09S 2.60 1.44 3.24 -11.90 - 8.59 4,29 - 1.40 2.90 1.59 F(H) -1.66 1.30 222 -1.16 038 212 204 - P16) 260 - .66 - .50 030 054 - 42 270 ON AUIS UW DH OX nw OV CON AU OD H Ox Dn ny} | -~l NOU EU EOF AAUP DH Of { iY) XO OVI & Ul rH Ow 7200 3.59 10.62 1.27 2.60 4.88 717 4.36 19.67 6.39 2.87 1.53 8.32 4.56 3233 1.60 1.34 53 331 375 2.68 264 065 5.83 7.76 6.65 266 2,32 1.44 2225 263 5.17 1.59 1.44 2.32 76 Fe 748 337 12,64 2.34 1.90 5.26 8.40 5-55 2.35 7 AT 1.52 022 9.58 4,11 4.33 1.73 2.73 1.50 214 3 204 343 13 80 3039 6.40 8.87 7.04 1.66 2.18 2.23 4.24 046 4.78 095 - 6,27 083 266 3270 089 F(H) 026 036 «50 206 tw ° BI oO e Wt Ol bh O22 he) e tl fn! e NAN OVLUTFPUNDH OF NOU PUDH ORD WYHOxX, FUWmHOrH Ode Mot STAT EUW YH Of lFol 265 1.45 1.31 1.41 264 264 1,09 1.01 2.06 66 065 063 266 11.15 589 6.97 1.80 TeD4 262 MWe fe * ¢ @ NPE AON oOorFroodon 8.38 8.81 6.87 3-21 3.01 77 228 2.00 1.10 337 1.48 213 1.54 242 2.42 1.12 82 224 1.56 1.71 11.06 AAD 6.70 057 7.72 o51 (2.352 016 208 1.78 1.17 1.97 2.07 1.54 49 1.50 9.62 8.60 4.98 2.60 2.25 1.64 2.61 2.02 F(H) 022 064 216 022 12 68 | NOW fF OLN eH Or ul ON vil NOAUEUND KOS NOU FUND HONS ul co} OUI & lH Ope AU UID HOR WI e nv) 1oN) 6.63 6.10 264 3.41 3.87 2.14 1.19 6.50 3057 2.69 064. 2.70 2.62 3.34 4,89 1.15 78 Fo -12.16 ~- 2.94 1.89 6.61 4.11 - 2,44 - 1.81 262 - 1.14 9.86 2.06 345 208 - 095 2.53 1.19 1.01 0S2 - ela 1.11 4.94 240 ~ 2.17 ~ 3,22 236 - 1.86 2.32 2.61 5-73 ~ 2.91 «70 - .92 341 F(H) ” 238 74 WN e WN ° 1S, \O => Wf Ul pp eH O93 i | na u 2m UNDKFO”Y UFrUnNrHO ff, rol UPUMDHF OMS ULM FO = il OM FU DH OS ol 2.42 61 2.40 3.16 263 261 1.37 095 1.26 2.25 1.76 059 1.40 1.52 5.12 341 13.08 2.52 3.83 2.18 3275 5 34 1,20 4.71 3.18 79 bi @ ft @eae Fg 3.42 1.27 1.62 66 244 1.46 10 1.48 1.16 12 264 1.22 2.00 2.69 098 76 228 1.23 057 3.31 039 4.38 2.17 4.73 1.71 023 4.12 5-90 050 4.07 3.26 10 2015 > > int fr ATEUDH OF AUN EFUDH OME f> OV Ul fl 9 FO 390 FUME OS® UFUNDrH Ow. UPUD HOPS [Fol 261 7.54 1.42 262 2.78 2.76 2.42 1.82 5.26 1.30 1,01 3.37 1.52 ~60 3 C7 60 261 059 61 «52 80 Fe 1.22 6.63 237 1.62 3.76 1.00 1.5 3.58 036 1.25 5.05 268 67 63 252 3.74 2,04 2.17 005 1.72 5.30 1.54 72 019 210 070 1.78 1.52 039 1.88 295 038 2.85 051 053 2.28 F(H) Aid al - p—) e a e | e ul 4 MS WHOHOSD NHOWDOS UWNHO uw ul oil fol Wi { 0 UDR O22 rUNMrF OF UNEHO WI 83] Ww OV wn Hox! Fup Fr O 261 059 1.13 1.76 2.14 ft f> f r woo LO OVW 61 2.09 1.36 81 °19 1.21 224 074 053 1.40 036 017 266 056 3219 1.17 2.01 1.74 5353 058 261 ol2 1.00 1.31 1.83 039 87 AAD 3.58 2.17 3.68 1.13 034 2.97 1.52 1.27 OI WN HM O= UNM HOM wn Col Ww ° | © ul e Or NH OX | pal nO = WON AU LIP HAsO OO COON OVUT EU fo Ho [Fol 261 1,50 14,26 33.12 29.71 4.30 2,20 3.21 66 LAT 099 12,07 4.10 31.24 2.78 2.55 4,08 1,81 6.50 82 Fe 024 1.74 210 ell - 2.18 1.58 048 1.55 - .54 F(H) 2.06 1,22 226 1.50 212 068 046 14 220 o1S ia] ol ® I HI UI ODA EU EOS OND PUND HS OONAU FUNDS OONAU FUDH ON 9.83 9.34 2.78 94 65 6.42 ALAS 1.55 1.33 3.49 9.61 6.17 5095 3.12 83 Fe ~10.87 19.48 - 9,24 087 3.83 12.09 1.44 - 4.14 5.20 9.23 7275 1.32 6.42 8.77 055 3.72 1.84 11.52 11,05 1.50 - 14 -/2 - 6.02 4.73 095 6.14 - 1,03 1.40 ~10.54 - 6.65 4.54 F(H) -1.84 -3 ot 12 268 - 70 218 224 032 048 | “172 oO ATEPONHMO NAW EOUDFH ro) | ed WOON OUFUND EI UN He LUD Hr 4.37 1.36 1.52 4.36 65 1.23 2.20 4,27 3.50 3.12 2.43 1,01 1.02 1.34 1.93 1.37 1.40 1.67 1.19 2.15 1.15 66 78 9.76 085 10.34 7.32 6.17 5.37 1.92 791 84 4,02 1.69 261 4.88 259 2.72 3023 3.49 3677 5.45 87 1.86 077 1.10 1.62 4,06 1.21 2.352 2.57 1.77 42 2.86 1.88 ool 250 8.65 1.67 8.49 6.20 4.88 4.33 2.15 7 tA 240 F(H) 290 2.38 - »f2 = .26 WO ONAL M HAS prof f- WOONKHAUFWDOHOeE OON AU FOUDHSA rol ul CON CuI & OD He 7.72 6.15 11.53 1.03 11.79 5233 6.64 2.17 1.86 1.38 4.62 7.89 6.01 065 1.80 3.289 065 2.12 17.4 9.74 4,83 2.64 1,24 2.66 1.61 099 14.78 1.06 6.12 265 3.87 85 - 9.13 5.11 10.56 1.80 11.68 ~ 5.02 7.08 ~ 1.25 1.48 - 2.46 8.09 752 86 - 2.12 Sell ~” 36 ~20.06 -10.68 1.351 ~ 4,39 3.04 1.80 - 057 - 1.62 - 2.67 092 ~ 5.75 5.99 078 =~ 4.40 F(H) 10 - .68 <a d 41.32 024 - .92 - .68 -1.72 2 24 -~ .64 204 nol =~] SOUPS ON NDUEUVDE XS rt en) °° rl 0 FUP HS UU FUNDS AN EUV NOUPUWDE nol e 4 ro) e i) e j e 86 rey Cc eos @ © @ @ NH OOH SOLO MN OUT 10 WW Ht po W191 OVW J 87 lr. Fo F(H) sy 1 4,42 3.17 40 2 12.48 11.94 - 30 3 5.88 - 4,84 220 A 2.59 - 2,47 5 2.26 oAL 8 1.81 1.67 9 «63 = 22 329 1 2.74 2.44 78 3 11.91 11.10 e 452 4 262 - 2,58 5 9.09 8.60 6 2.94 ~ 2.12 7 3.94 343 8 2.53 2.09 3352 | 1 251 2.20 052 2 9.32 - 9.51 036 i) 11.50 12.26 238 A 4,29 4.79 5 2.61 1.80 6 63 - 1.20 7 1.87 1.14 8 2.92 - 1,21 9 3.51 423 ZAQ 1 261 32 2 2.25 - 3.72 oth 3 2.02 3.54 4 3.06 3.48 5 2.85 - 3.39 6 1.44 - 1.98 8 2.43 - 4,06 9 63 3.24 vil x l ra) CON NOU FO DH WW = OV SIOUEUND ERP ON OU EUNDE dl fee) AN FUME AO PUD He ry Qo ° e oJ e * e QUI UT UT UT UT O He ee Pei OAAVYANYAD On ~ 2.91 1.43 4.32 4.65 1.50 264 1.58 261 4.76 264 092 88 ry] ) © e NVUIW NAO O HH OF OW AU . Hue UI e ~ 1.72 3.50 - ,02 - AL74 3.77 2.41 65 342 1.01 ~10.68 ” 5.73 2.52 1.90 2.41 ~ 1.47 - 4,52 - 2.17 ~ 1.90 2.04 - 4.34 2.32 269 208 5.39 - 2,69 | 052 1.26 1.23 F(H) ll ° =) Oo e > ON AU FUND RSS FUND Re UFO MRA tof ) = ul ONAU EU DHS ON aOUFU DH ol 1.94 054 262 064 264 059 5.75 2.68 1.83 1.83 2.05 3476 4,84 2.69 1.42 5255 60 794 4.61 1.25 4,60 3.01 1.62 3.54 10.32 2.07 3055 059 89 Fe 1.35 2.75 1.36 1.34 beteeee ul e Ui rs f te Cd (o) as F(H) CON OUI WW PAO | rs] ae] OTF un Se NAUEUNDHRS ON DU EUD HOS el co ON FUVHS Nau PUNDH lFol 059 4.16 1.44 1.35 260 3255 261 1.80 ASA 3677 4,12 1.40 6.75 1.95 1.75 5.38 2.56 4,06 90 5.08 1.21 72 4.10 6.08 4,62 292 719 1.92 1.86 1.10 204 5-35 7.61 5.83 1.95 23 1.94 eel 3.81 1.24 ~ 4.49 289 203 273 54 3268 035 1,28 F(H) 8 © SN QAUIEFU MRO ATDOAWUFU DHS NNO EUD HO WHO FUND Hee UW PW De 262 62 1,29 2.81 1.39 260 260 61 259 404 3.48 1.79 .60 261 1.69 390 3.34 2.69 1.30 3-51 61 91 1,66 240 296 279 1.62 - 1.05 2.51 eD7 edl 077 046 039 2.71 5.56 O07 42 1.34 214 1.31 053 4.58 2.30 1.60 1.21 054 1,24 3.31 1.57 1.75 1.06 2.44 1.86 al => UI AOU EWVHP ANAU EOD NOU EUW eH 57 PY Ww ies) l ul Xe) PUNE FP UU EAS AU EU 2.85 1.38 2.75 1.34 92 @#aotbts 3.58 2.71 77 2.23 1.46 2.15 366 1.25 2.10 019 4,16 1.86 ae) 1.47 075 3.97 4.49 2.95 1.88 0.00 1.14 1.27 oft 2.78 1.27 2.27 3674 «87 1.41 261 204 2.67 216 F(H) 61g 2 628 em FUNH at On} & WN FUDD ROS UNDE con) Ul WP Ho 66 aS &W PD eH ol 260 263 1.53 63 1,04 089 93 Fe 260 1.05 053 40 057 261 034 we) 1.55 1.62 2.81 1.92 1.40 209 «83 083 ef0 2.00 2.06 039 F(H) ee Se ae De Se LO. ill. 12. 13. 18. 19. 20. 94 REFERENCES F. Feist, Ber. 26, 747 (1893) D. T. Jones, J. Chem. Soc. 87, 1062 (1905) F. R. Goss, C. K. Ingold, and J. F. Thorpe, J. Chem. Soc. 123, 327 (1923) ———— M. G. Ettlinger, J. Am, Chem. Soc. 74, 5805 (1952) C. K. Ingold, J, Chem. Soc. 121, 2676 (1922) F. R. Goss, C. K. Ingold, and J. F. Thorpe, J. Chem. Soc. 125, 5342 (1925) ~——— F, Feist, Ann, 436, 125 (1924) F, Re Goss, ©. K. Ingold, and J. F. Thorpe, J. Chem. Soc. 127, 460 (1925) F. B, Thole and J. F. Thorpe, J. Chem. Soc. 99, 2187 (1911) and later papers F, C, Whitmore, Organic Chemistry, Ven Nostrand (1937) Pe 475 J. W. Baker, Tautomerism, Routledge & Sons (1934) p. 164 G. A. R. Kon and H. R. Nanji, J. Chem. Soc. 2557 (1932) D. Lloyd, T. C. Downie, and J. C. Speakman, Chem.& Ind. Feb. 20, 1954, p. 222 L. Long, Chem, Rev. £27, 437 (1940) J. T. Gragson and others, J. Am. Chem, Soc. 75, 3344 (1953) Professor J. D. Roberts, private communication I. Heilbron, Dictionary of Organic Compounds, Oxford, S» S61 (1953) D. P. Shoemaker, Thesis, CIT (1947) Chia-si Lu, Rev. Sci. Instrum. 14, 331 (1943) A. J. G. Wilson, Nature 150, 152 (1942) ele Les 256 24. 27, 28, 29, 30. 31. 32. SS-6 ote S56 56-6 S7. 95 W. A. Wooster, Textbook on Crystal Physics, Cambridge (1938) p. 207 BE. R. Howells, D. C. Phillips, and D. Rogers, Acta Cryst. fe) 210 (1950) wan? A. L. Patterson, Zeit. f. Krist. 90, 517 (1935) J. D. Morrison and J, M. Robertson, J. Chem. Soc. 980, 987, 993, 1001 (1949) J. C. Speakman, private communication D. Lloyd, T. C. Downie, and J. C. Speakman, Chem. & Ind. April 24, 1954 T. C. Downie and J. C. Speakman, Acta Cryst. 7, 649 (1954) E. W. Hughes, J. Am. Chem. Soc. 63, 1757 (1941) V. Schomaker and R. A. Pasternak, unoublisned description D. P. Shoemaker and others, J. Ams Chem. Soc. 72, 2578 (1950) D. W. J. Cruikshank, Acta Cryst. 2, 65 (1949) H. G Pfeiffer, Thesis, CIT (1949) L. Pauling, Nature of the Chemical Bond, Cornell University (1948) p. 171 T. H. Goodwin and C. M. Thomson, Acta Cryst. 7, 166 (1954) J. Trommel and J. M. Bijvoet, Acta Cryst. 7, 703 (1954) Professor V. Schomaker, private communication P. Vaughan and J. Donohue, Acta Cryst. 5 580 (1950) women 96 PROPOSITIONS 1. The structure suggested by Jones* for the tribasic . acid CgoHg0, is probably incorrect. A more satisfactory structure is proposed. *D. T. Jones, J. Chem. Soc. 87, 1062 (1905) — 2e Molten sulfur becomes viscous and darkens in color as its temperature is raised to about 230°.# It is proposed that a similar behavior may be observed at a lower temperature if the molten sulfur is bombarded with high energy electrons. *L., Pauling, General Chemistry, Freeman & Co. (1953) De 556 Se It is proposed that an interesting substance for structural analysis is the "two-dimensional" network of silicon atoms recently described.* A procedure is suggested. | tH, Kautsky and L. Haase, Zeit. f. Nat. 8b, 45 (1953) as reported in Curr. Transl. 2, No. 2-3 (1954) p. 1 4, Concerning the separation of Co(II) from Ni(II) by paper chromatography with a ketone-nydrochloric acid solvent, * it is proposed that: (a) the separation is of the molecular species NiClo*S6HOH and HoCoCly; and (b) the use of NH, is an aid in locating and 97 identifying ion concentrations. *T, V. Arden and others, Nature 162, 691 (1948) 5. A recent structure investigation* includes results which indicate that the thiocyanate ion is non-linear. It is proposed that this ion is shown linear by other experimental and theoretical considerations. *H. Scouloudi, Acta Cryst. 6, 651 (1953) 6. If a crystal shows anisotropic scattering in the direction a,, the correction for anisotropy to be superimposed on the isotropic temperature effect reduces to exp(qh,”), where q is a constant and hy is the index associated with a,.* - *This thesis, pe 33 7. The mechanism for the addition of ozone to carbon-carbon @ouble bonds is not yet clear. It is proposed that: (a) an. examination of the structures of crystalline ozonides* would give useful evidence; and (bp) the anomalous ozonolysis of Feist's acid** may aid in the interpretation of the ozone reaction. #R. Oriegee and G Lohaus, Chem. Ber. 86, 1 (1953) *#This thesis, pe 47 8. The abnormal reaction of ozone with Feist's acid suggests that: (a) a careful re-examination of the products of the hydrolysis of the ozonide of methylenecyclo- propane* might show the presence of some acetyl- i i q i | 'q if iH 98 acetaldehyde or related compounds; and . (b) the establishment of the structures of cyclo- propene derivatives** by ozonolysis is not reliable. #J. T. Gragson and others, J. Am. Chem. Soc. 75, 3344 (1953) weg, R. Nunn, J. Chem. Soc. 313 (1952) 9, A traveller at point P on the surface of the earth moves 100.00 km due south, then 100.00 km due east, and then 100.00 km due north to find himself again at point P. It is proposed that there is more than one such point from which the traveller mignt begin. 99 The following proposition replaces number six on Page 97, which is unsatisfactory: | It is proposed that the experimental evidence* for the stability of Ie°M compounds is related to the evidence for the effectiveness of M as a third body in the . recombination of iodine atoms*# and that some of the former evidence may be predicted qualitatively from the latter. *H. Benesi and J, Hildebrand, J. Am. Chem, Soc. Wl, 2708 (1949) *#Russell and Simons, Proc. Roy. Soc. A 217, 271 (1953)