.r,;1.. I a".II'. ' ...‘ :JI‘-\ ‘. unr‘ "L' '3‘ I‘I I sci. :éhl'hk-l” I Iz‘bz; .“."w 2" \EA; ‘- 'II u.‘ “I . I'j.‘ " ' J. '1‘" I. ”In. P_ ‘5‘“ . A .V 1" L "-I HUO'U-t . -,.A 1583' Nu"- -- Ni" b." I .. AI -'-'_ - - if!“ “I, ,A '59,?“ I',"..v.-. "l I j ‘ n1" " I; v; “II.~“‘~J;"’" ”I... u I 1": 5'4 - I vi‘ flag-*9“ n 199: h I» ‘h ': I _< "14):. J," at: V N u .C,‘ .I gum-ma :t‘ I» ' \ .n h I‘ll-I «wax; “ , 2:113; '.I;‘. L‘I'J'A I. vl 'AoA. Ir; _ AIG'IX‘T; \‘ . l Hit L: a It. (:3, . OVA)“ A ..~. «find! 2.21;. " ' S. usiv ' a. .4“,". ."v '. (' .. -‘.' “‘\~-)'Av" ‘3: MI“ \ '1’. II‘ .I-O v "‘ ‘ ‘.'."'l"l ‘ I. .‘I'V‘ - \'...I" “-3.- 'q-‘ . :TIK‘KHE: ‘ln . I. «11,3 ‘ ' .m ' . : «.3 ”'32. :2wa d ‘63}; §2;;.;.-.-:;3 I A n’A‘» a“ fix; .. +3: $9131 .. Ct“ PfJggv .... r -- mew—+- LIBRA R Y ? Michigan State University THES'S This is to certify that the thesis entitled Spectroscopic Studies of Alkali Metal Complexes in Non-Aqueous Media presented by Adamantia Rokofilou-Hourdakis has been accepted towards fulfillment of the requirements for Ph.D. degree in Chemistg fimfifiu/ Major professor Date film/76 0-7639 ' 113M , j fink-‘l ‘, '3. III . ‘ ' ~ . -I‘IIII/ ~ l ‘I "A ' OVERDUE FINES: 25¢ per day per it. RETUMIM LIBRARY MATERI£§3 Place in book return to my charge from circulation new SPECTROSCOPIC STUDIES OF ALKALI METAL COMPLEXES IN NON-AQUEOUS MEDIA By Adamantia Rokofilou-Hourdakis A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1978 6//5A3;L2 ABSTRACT SPECTROSCOPIC STUDIES OF ALKALI METAL COMPLEXES IN NON-AQUEOUS MEDIA By Adamantia Rokofilou—Hourdakis The complexation reaction of the dilactam of cryptand C222 with lithium, sodium and cesium ions in water and in several nonaqueous solvents was studied by using 7Li, 23Na, 133Cs and 130 NMR technique and far infrared measurements. Alkali metal ion chemical shifts were determined as a function of dilactam/metal ion mole ratios. There is no indication of a Li+-C222D complex formed in water, dimethyl sulfoxide, dimethylformamide and methanol while in formamide and acetone there is indication of a rather weak complex. Formation constants of lithium-C222D complexes were determined in acetonitrile, tetrahydrofuran, pyridine and nitromethane. The values obtained were: log KACN = 3.23 I 0.07, log KTHF = 3.12 i 0.09. log KPy = 2.64 I 0.10, and log KNM > 4. Sodium ion forms stronger complexes with C222D since its size is closer to the dimension of the dilactam cavity than that of the lithium ion. In methanol. dimethyl- formamide, nitromethane. acetonitrile and pyridine solutions log K > 4 has been found. The large cesium ion does not fit conveniently into the C222D cavity. The Cs+-0222D complex is rather weak in nitromethane. acetonitrile and pyridine solutions while there is no indication of complex formation in dimethylformamide and in aqueous solutions. In order to Adamantia Rokofilou-Hourdakis calculate the formation constants of the Cs+—0222D complex in the above solvents, the ion pairing which is particularly extensive in the case of pyridine, was also taken into account. The values obtained were: log K = 1.79 i 0.03, NM log KACN = 1.70 i 0.08 and log KPy = 2.86 i 0.01. The limiting chemical shifts of the complexed lithium and sodium ion are solvent-dependent, indicating incomplete insulation of the cation from the solvent. while the large difference in the limiting chemical shifts of the complexed cesium ion is a good indication that the metal ion remains exposed to the solvent. Carbon—13 NMR Spectra of the Li+-0222D, Na+—C222D and KI-C222D complexes in nitromethane show that the complexes are of the inclusive type with the metal ion inside the ligand cavity. Formation constants for the Cs+—C222D and Li+-0222D complexes in pyridine, nitromethane and acetonitrile were calculated at various temperatures and used to obtain the thermodynamic quantities AH and AS for the complexation reaction. The results for the Cs+ with 0222D reaction were: AH = —2.18 i 0.09 kcal/mole, AS = 5.69 i 0.33 cal/mole-OK in pyridine, AH = -o.67 : 0.10 kcal/mole, AS = 5.81 i 0.33 in nitromethane and AH = —l.85 : 0.26, AS = 1.37 i 0.93 in acetonitrile. The complexation of Li+ with 0222D is endothermic because the small lithium ion is strongly solvated and the energy required for the desolvation is not replaced in the Adamantia Rokofilou—Hourdakis complexation step. The values obtained are: AH = 6.10 i 0.16 kcal/mole, AS = 38.4 i 0.6 cal/mole-OK in nitromethane and AH = 1.58 i 0.22 kcal/mole, As = 21.7 i 0.9 cal/mole-OK in acetonitrile. The complexing ability of 2,2'-Bipyridine with lithium, sodium and cesium ions in various solvents was studied by using 7Li, 23Na, 133Cs and 13C NMR technique. The metal ion chemical shifts were determined as a function of ligand/metal mole ratios. There is no Li+—2,2'-bipyridine complex formed in tetrahydrofuran, methanol and formamide while in propylene carbonate there is a weak complex formed. In nitromethane a strong complex of a 1:2 stoichiometry is formed as both the 7Li and the 13C data indicate. The Na+-2,2'-bipyridine complex seems rather weak in tetrahydrofuran, propylene carbonate and acetonitrile, while there is no indication of complexation in dimethyl sulfoxide, dimethylformamide and methanol. In nitromethane there is a complex formed but we cannot unambiguously conclude its stoichiometry. The complexation of 2,2'-bipyridine with Cs+ was studied in nitromethane, dimethylformamide, pyridine. propylene carbonate, acetonitrile and methanol. In none of the above solvents is there an indication of a complex formed. ACKNOWLEDGMENTS The author wishes to thank Professor Alexander I. Popov for his guidance, counseling, and friendship throughout this study. Professor Andrew Timnick is acknowledged for his many helpful suggestions as second reader. Gratitude is also extended to the Department of Chemistry, Michigan State University and the National Science Foundation for financial support. I would like to thank all the members of the laboratory of Dr. A. I. Popov for their friendship and encouragement during our association, and Mr. Frank Bennis and Mr. Wayne Burkhardt for their efforts in keeping the NMR spectrometer in operating condition. Special thanks to our friends in the U.S.A., who contributed to render our stay in this country a pleasant and unique experience. ii TABLE OF CONTENTS Chapter Page LIST OF TABLES........................................ LIST OF FIGURES........................................... vii I. HISTORICAL REVIEW ................................. 1 A. Complexation of Alkali and Alkaline Earth Metal Ions by Cryptands ...................... l B. Study of the Thermodynamic Parameters of the Complexation of Crowns and Cryptands.......... 13 Conclusion ........................................ 19 II. EXPERIMENTAL PART ................................. 21 A. Salts ........................................ 21 B. Ligands ...................................... 21 C. Solvents ..................................... 22 D. Sample Preparation ........................... 24 E. Instrumental Measurements .................... 24 1. Lithium-7, Sodium-23 and Cesium-133 NMR.. 24 2. Carbon-13 NMR ........................... 25 3. Far-Infrared Spectra .................... 26 4. Data Handling ........................... 27 III. SPECTROSCOPIC STUDIES OF COMPLEXATION OF ALKALI METAL IONS. WITH THE DILACTAM OF C222.............. 28 A. Lithium-7, Sodium-23 and Cesium-133 NMR Study of C222D Complexes in Various Solvents ........ 8 B. PH Dependence of the 23Na Chemical Shift ...... 48 iii & TABLE OF CONTENTS (Continued) Chapter Page IV. VI. C. pKa Determination of 0222D .................... 52 D. Far-Infrared Study of Lithium and Sodium Complexes with C222D in Nonaqueous Solvents ... 54 E. Carbon-l3 NMR Study of Lithium, Sodium and Potassium Complexes with 0222D in Nitromethane IUCCCIIUOOIICII'DIIIIIDIIIIIIIIIII 5? STUDY OF THE THERMODYNAMICS OF THE COMPLEXATION REACTION OF CZZZD WITH LITHIUM AND CESIUM IONS IN VARIOUS SOLVENTS OI....IIIOOOICOIOOCICCOOIIIIODIICI64 Introduction 0......IUOIIIIOIIUCIIIOOO....IIIIOU... 61+ Results and Discussion ............................ 65 SPECTROSCOPIC STUDIES OF COMPLEXATION OF ALKALI METAL IONS, WITH 2.2'-BIPYRIDINE IN NONAQUEOUS SOLVENTS.. 88 Introduction ...................................... 88 Results and Discussion ............................ 90 APPENDICES APPENDIX I ........................................127 APPENDIX II .......................................132 LITERATURE CITED ......UIIDICCIIIIOCUOIIIOCOOIOIIIOI.IIClZO iv fl Table 1 10 11 12 LIST OF TABLES Ionic Radii of Metal Cations. Approximate Cavity Radius and Number of Binding Sites of Ligands ...... Thermodynamic Quantities of Alkali-Cryptate Compl xation Reaction Measured by Calorimetry at 25Cinwater IIIIIIIIIIOCIOI..IOQIIUIIIIICIIIIDI Key Solvent Properties and Correction for Magnetic Susceptibility on DA—60 ................... Lithium—7 Chemical Shifts at Various M018 Ratios, [C222Dl/[Li l, in Nitromethane at 33 i 2 C ......... Sodium-23 Chemical Shifts at Various Mole Ratios, [0222Dl/[Na+l, in Various Solvents, at 25 i 2 C .... Cesium-133 Chemical Shifts at Various Mole Ra ios, [0222Dl/[0s+], in Various Solvents, at 25 1 2 C .... Formation Constants of CZgZ-Dilactam Complexes in Nonaqueous Solvents at 25 C ........................ pH Dependence of 23Na Chemical Shift for Aqueous Solutions of Sodium Chloride ....................... Li-7 Chemical Shifts of LiClOu (0.010 M) in the Presence of 0222D in CH CN at Various Temperatures III'CICICI....IDUICOI.....IIIOIOIIOICCI Li-7 Chemical Shifts of LiClOu (0.010 M) in the Presence of C222D in CH3N02 at Various Temperatures IOIIIOOIIIIIDIIIIII.IIIIIIOOIIIIOIIJIII Li-7 Chemical Shifts of LiCth (0.020 M) in the Presence of 0222D in Pyridine at Various Temperatures .....ICICCDOII.IIIDDDIOOOICIOIOOOIIIIII Formation Constants for the Complexation of LiClOu By 0222D in CH3N02 and CH3CN at Various Temperatures I.ICU...CICOOIOII'.IIOIIIOOIIIIIIOI'II. Page 6 18 23 30 31 50 67 69 7O 75 LIST OF TABLES (Continued) Table Page 13 14 15 16 17 18 19 20 21 22 23 24 Cs— 133 Chemical Shifts of CsBPhu (0.015 M) in the Presence of C222D in Pyridine at Var1ous Temperatures OICOIOOIII.OIIIOII..IIOIIIIOIO.IOIICOII 79 03-133 Chemical Shifts of CSSCN (0.010 M) in the Presence of C222D in CH N0 at Various Temperatures IIIQOIIIIIIICZIOIIIIIcicle-0|.Ilolotllll 80 Mole- -Ratio-Temperature Data for the Chemical Shift of CsSCN (0.01 M) in the Presence of C222D in III-I0.0I00......UCUIIIIII..IIIOODIIIIIOIOIUIO 81 C 3 N Formation Constants for the Complexation of Cs+ by 0222D in Py, CH NO2 and CH CN at Various Temperatures IIIIIIOIOOOIOIIIIIIIIIIIIOIIODOOIIOIU'I 82 Thermodynamic Quantities for the Complexation of Li+ and Cs+ by C222D in Various Solvents ............... 87 Mole Ratio Study of 2, 2' -Bipyridine Complexes with Lithium in Various Solvents by Li- 7 NMR at 25°C..... 91 13C Chemical Shifts (ppm) for 2, 2' —Bipyridine (0.20 M) with LiClOu in Nitromethane ........................ 97 Mole Ratio Study of 2, 2' -Bipyridine Complexeso with NaBPhu in Various Solvents by Na- 23 NMR at 25°C .... 99 130 Chemical Shifts (ppm) for 2, 2' -Bipyridine (0.20 M) with NaBPhu 1n Nitromethane ............... 107 Mole Ratio Study of 2, 2' -Bipyridine Complexes with Cesium in Various Solvents by Cs— 133 NMR at 25C... 111 Mole Ratio Study of 4 Mh'-Dimethyl, 2, 2' O-Bipyridine Complexesowith0.150MNaBPh4 in CH3NO by Na- 23 NMRa-t 25C IIII.C.IIICCOCIOOOIIIIIOIIIIIICICCIIOIII 115 Mole Ratio Study of 2, 2' —Biquinoline Complexes with 0. 075? M NaBPh# in CHBNO2 and DMSO by Na— 23 NMR 25 a CIOIICIOIIIIIIIOIOOIIII‘IOIIIOOIIIOIIIIIOOOOI 118 vi LIST OF FIGURES Figure l DibenZO-la-CI‘OWn-6.......”......u................... 2 a) Cryptand, C. b) The dilactam of 0222........... 3 Exo—exo, endo—endo and exo-endo conformation of C222 CryptandollClIIIIlIllIIIIIIIIIIIOIIUIIIIOIIIIOII h Intramolecular cation exchange in mononuclear complexes of a macrotricyclic cryptand............... 5 Newly synthesized macrotricyclic cryptand............ 6 AnioncryptateilflilIllOIDCQOOIIII.IIIIIICIII‘IOIIOIO. 7 Lithium-7 chemical shift as a function of 0222D/Li+ mole ratio in dimethyl sulfoxide, methanol, aqueous, formamide, dimethylformamide, and acetone solutions.. 8 Lithium—7 chemical shifts as a function of CZZZD/Li+ mole ratio in acetonitrile. tetrahydrofuran, nitromethane and pyridine solutions.................. 9 Sodium—23 chemical shifts as a function of 0222D/Na+ mole ratio in nitromethane, methanol, dimethyl- formamide. acetonitrile, aqueous and pyridine SolutionSOIIOOIIIOIIIO.IOIOIIOOOICIICIIOOIOIIIIOIOIIl lO Cesium-133 chemical shifts as a function of 0222D/Cs+ mole ratio in nitromethane, dimethylformamide, aqueous and pyridine solutions...................... 11 Cesium-133 chemical shifts as a function of C222D/Cs+ mole ratio in acetonitrile.......................... 12 Changes in Na+ resonance as the relative amounts of free and complexed Na vary with changing pH for 0.05 M aqueous NaCl with [C2221/[Na+] = 1.0......... 13 Titration curve for the titration of 25.0 ml of 4 x 10-3 13 aqueous 022213 with 0.05 MHCl............ 14 Carbon-13 spectra of: a) 0.10 M 0222D in CH N02; b) 0.10 M C222D with 0.2 M LiClOu in CH3N02......... 15 Carbon-13 spectrum of 0.25 M 0222D in CH N0 (Bruker 180 NMR spectrometer)..............g........ vii Page 10 ll 12 35 36 41 43 44 51 53 58 59 LIST OF FIGURES (Continued) Figure Page 16 16A. 17 18 19 20 21 22 23 24 25 26 27 28 29 Carbon- 13 spectra of: a) 0.05 M C222D with 0. 06 M NaBu in CH N02 , b) 0.05 M C222D with O 10 M }{PF61+ inCH 3R022...‘.|CI:....IIIII.....IOI:...—IIIIII.III 61 IR spectra of the carbonyl absorption of: a) 0222D in nujol: b) Na+ -0222D in uiol c) [Li+]/[C222D] = = l. 2.0 in CHBNO -d) [K+]/[0222D 6 in CHBNOZ........ 62 Lithium—7 chemical shifts as a function of 0222D/Li+ mole ratio in acetonitrile at various temperatures... 71 Lithium—7 chemical shifts as a function of 0222D/Li+ mole ratio in nitromethane at various temperatures... 72 Lithium-7 chemical shifts as a function of CZZZD/Li+ mole ratio in pyridine at various temperatures....... 73 A plot of 1n K X§ 1/T for the complexation reactions of LiT with C222D in acetonitrile and nitromethane... 76 Cesium-133 chemical shifts as a function of 0222D/Cs+ mole ratio in pyridine at various temperatures....... Cesium-133 chemical shifts as a function of CZZZD/Cs+ mole ratio in nitromethane at various temperatures.... 84 Cesium-133 chemical shifts as a function of 0222D/Cs+ mole ratio in acetonitrile at various temperatures.... 85 A plot of in K gs l/T for the complexation reactions of Cs+ with 0222D in pyridine. acetonitrile and nitromethane....IOIlCOOIIIOIOIIOIIIIQOIOIIIIOIIIOIIII 86 Lithium-7 chemical shift as a function of 2,2' BP/Li+ mole ratio in THF, MeOH, DMF, and PC solutions....... 93 Lithium-7 chemical shift as a function of 2,2' BP/Li+ mole ratio in nitromethane solutions................. 94 Carbon- 13 chemical shifts as a function of Li +/2, 2' BP mole ratio in nitromethane solutions. ................ Sodium-23 chemical shift as a function of 2,2 BP/Na+ mole ratio in DMF, CHBOH, THF, PC, CH CN. and DMSO solutionSOIC'CCOCIIIUI....IIIOUCOOI...'IICIICOIOUIOD 102 Sodium-23 chemical shift as a function of 2,2' BP/Na+ mole ratio in nitromethane solutions................. 10b viii LIST OF FIGURES (Continued) Figure Page 30 Line-width change of the sodium-23 resonance as a function of mole ratio in CH N02. PC, THF and CH CN SolutionSlIOI.IICIIOIOICI.IOIIOOIIOOIOI'...I.I..IIIII 105 31 Carbon-l3 chemical shifts as a function of Na+/2,2' BP mole ratio in nitromethane solutions................. 32 Cesium—133 chemical shift as a function of 2,2 BP/Cs+ mole ratio in CH N02, CHBOH, Py, PC, mlF and CHBCN Solutionsflfl......IIIIOIICC.IIIOIIIIOI.‘ 113 33 Sodium-23 chemical shift as a function of 4.h' dm,2,2' BP/Na+ mole ratio in nitromethane SolutionsIlIUOOO.IIIIOII.OIIIIIIICI......DICIIIOOOOCI 117 3b 4- Sodium-23 chemical shift as a function of 2,2 BQ/Na mole ratio in nitromethane solutions................. 119 ix .- 5W4; A WW I ;' Although alkaii and alkaline earth metal cations play I role Cf great impor:en:e bctfi in chemistry and in biology. their coordinnt;an chemistry has vainly been duveloping in recent vents wi,h the adVont C‘ ~1~nral (1) or synthetic macrocyclic and necrogslytyclic “'.w§:. which are capah; of forming strong .‘H,Le e? x.’u ~ : eve :ctal ions. Nacxc23cli: pclvethers ': ' _ n.” éii‘rfi, synthesized in 1947 by Ecuersen (2.? MET: :?2 --.3. agents 'to appear. a fgj;aqi These macrocycifig Ligands Lents.“ leu:-ar. 1W“ dimensional ‘CHAPTER I > caVLty. the d'aaeta' n.f ~“ich es~ {a 'arLed by :.ung:nc the HISTORICAL REVIEW ‘ number of methyluwe grrzL I"? U! ..r‘* -?ff#:: .1 the ring. ‘9' Certain Cyalic {Oigethc*i ac? :rEv strorglv tlri carticular n5. ' . _ . ' -%F,; llktll and a.hu]Lwe»~ertn «at; _*nu h“: vaiqc*1:s.' bind $5”. $8” an. or warn of then! ': in frefcrfifiao t‘ '&9 ether- ,2 ‘ififi gorigs {n.43, an ;arlmetorr vh31h Lnxiuence to. liloetixitv nLi binding preperties -“ «a r “vvles .EC-1Jx 1h. typo(s) and nunber of binding -;'on :n :r« ring. the :.'ro13tivo 013.3 of the ion and the macrn<;yclic cavity. the lelical placement of the binding sites.stor1c hindrance flfil the ring. the solvent and extent 0: solvataar of the ion ‘Jhdin. lites. and the electrical charge of the Lou. : My momma. him and coworkers (6 to: Wand ‘ jg!»- ot canal-3“”. sun-poumumu " :wmm um.mu “About”; magi» A. Complexation of Alkali and Alkaline Earth Metal Ions By Cryptands Although alkali and alkaline earth metal cations play a role of great importance both in chemistry and in biology, their coordination chemistry has mainly been developing in recent years with the advent of natural (1) or synthetic macrocyclic and macropolycyclic ligands, which are capable of forming strong complexes with the above metal ions. Macrocyclic polyethers or "crown" ethers, synthesized in 1967 by Pedersen (2,3) were the first such complexing agents to appear. A typical crown is shown in Figure 1. These macrocyclic ligands contain a molecular, two-dimensional cavity, the diameter of which can be varied by changing the number of methylene groups and/or ether oxygens in the ring. Certain cyclic polyethers not only strongly bind particular alkali and alkaline earth metal ions but selectively bind one or more of these ions in preference to the others in each series (U,5). The parameters which influence the selectivity and binding properties of macrocycles include the type(s) and number of binding sites in the ring. the relative sizes of the ion and the macrocyclic cavity. the physical placement of the binding sites, steric hindrance in the ring, the solvent and extent of solvation of the ion and binding sites. and the electrical charge of the ion. Shortly thereafter, Lehn and coworkers (6—10) introduced a new class of complexing agents, diaza-polyoxamacrobicycles called "cryptands", which contain tri-dimensional molecular l Figure 1. Dibenzo—lB-crown—6. The number 6 refers to the total number of oxygens and 18 to the total number of atoms in the polyether ring. cavities (Figure 2). The size of the cavity can be varied by changing the length of the ether bridges. In general. the selectivity of complexation and stability of complexes are several orders of magnitude greater than that of the crown ethers with the same number of oxygen atoms in the ring. It has been suggested by Lehn that the term cryptand refer to the ligand and cryptate to the complex. In macrobicyclic complexes. the cryptand may exist in three forms differing by the configuration of the bridgehead nitrogens: exo—exo (x-x), exo-endo (x-n) and endo-endo (n—n) (Figure 3). These forms can easily interconvert by nitrogen inversion. Although it is not known in which conformation the free ligand exists in solution, the endo-endo form should be strongly favored in the complex since it allows both nitrogen atoms to participate in the complexation interactions. The crystal structure deter- mination of the ligand C222 and of several cryptates (ll-15) showed that the cation was indeed included in the center of the three-dimensional cavity of the macrocycle which was in the endo-endo form. exo-exo endo-endo ‘ exo-endo Figure 3. Exo-exo. endo-endo and exo-endo conformation of 222 cryptand. fl 0 o m Q0” C) U n Cf/F-—1\E) Ro’\o/ju ‘3 CK\L__d//C)‘\\//’L<) Figure 2. a) Cryptand, C. C-211 m = O, n = l C-322 m = l, n = 2 C—221 m = l, n = O C-332 m = 2, n = l C-222 m = n = l C—333 m = n = 2 c-22c8 m = 1 (third bridge = —(CH2)8-) b) The dilactam of c—222. 5 Cryptates of alkali and alkaline earth cations, show cavity selectivity, the preferred cation being that whose size most closely fits the ligand cavity (Table l). The cryptands C211, C221 and C222 thus complex preferentially Li+, Na+ and K+ respectively. Formation constants of alkali- macrocyclic ligand complexes have been obtained (16—18) potentiometrically in aqueous and methanolic solutions. The conclusions that the authors draw from their study is that ligands of the "rigid" type like 0211, 0221 and 0222 present a stability peak for the optimal cation. The ligands of the "flexible" type on the other hand as 0322, 0332 and 0333 which contain large, adjustable cavities, show plateau selectivity. In other words, the ligand shows large selectivity for K+/Na+ but cannot differentiate between K+, Rb+ and 03+. Control over alkaline earth-alkali cation selectivity may be achieved by increasing ligand thickness or change in the number of binding sites (18,19). The authors investigated 2+ in methanol the complexation selectivity for Na+, K+ and Ba and water solutions. The effect of decreasing the number of binding sites is clearly shown by comparing cryptands 0222 and 02208. While the first has a complexation selectivity Ba2+/K+ m 10h, the latter compound with two oxygen binding 2 for the same ratio. sites less, gives < 10' By replacing also the oxygen binding sites in cryptands 0211, 0221, 0222 by nitrogen or sulfur, they found (20,21) that as the number of oxygen sites decreases the stability and selectivity of the alkali and alkaline earth cation Table l. Ionic Radii of Metal Cations. Approximate Cavity Radius and Number of Binding Sites of Ligands. Cation Ionic Ligand Cavity No. of Radius Radius Binding ri(A)a (A)b Sites Li+ 0.78 0211 0.8 6 Na+ 0.98 0221 1.1 7 K+ 1.33 0222 1.0 8 Rb+ 1.09 0322 1.8 9 0s+ 1.65 0332 2.1 10 Mg2+ 0.78 0333 2.0 11 0a2+ 1.06 Sr2+ 1.27 Ba2+ 1.03 aPauling values. bMeasured on Corey-Pauling—Kaltun molecular models. complexes decrease rapidly while they increase for transition metals and metals of group IB and IIB. An especially interesting feature is that the polyamine cryptands provide a means of trapping transition metal cations inside the molecular cavity imposing thus coordination geometries and may modify the spectral and redox properties of the cations. The kinetics of the complexation by cryptands have been studied by a variety of techniques. Lehn gt g1. (22) studied the kinetics of the complexation by temperature variations in proton nmr measurements of K+-0222, Na+-0222 and Tl+-0222 cryptates in D20 solutions of cation exchange. They concluded that the mechanism is a dissociation-complexation process rather than a bimolecular process. The exchange becomes slower as the stability of the cryptate increases. The symmetrical splitting caused by Tl+-H spin-spin coupling indicates that the ion is in the center of the molecular cavity. Dye and coworkers (23) studied the exchange rates of sodium- 0222 cryptate in ethylenediamine by using the sodium-23 nmr technique. They found an activation energy of 12.2 i 1.1 kcal-mol'l for the dissociation of the complex. The rate of dissociation of the complex is also similar to that found by Lehn gt g1. (22) for aqueous solutions. Kintzinger and Lehn (24) used double probe 130 and 23Na nmr studies to obtain correlation times and 23Na quadrupolar coupling constants for sodium cryptates. They found that the 23Na nuclear quadrupole coupling constant decreased with an increasing number of oxygen atoms in the ligand. The chemical shift (referred to a 0.25 M aqueous NaCl solution as external reference), has values of -1l.l5 ppm for Na+-0211, +0.25 ppm for Na+—0221, and +11.45 ppm for Na+-0222. The line widths at half height were 132 j 3, 06 i 2 and 29 j 1 Hz respectively. Complexation studies were extended by Cahen £3 31. (25) to lithium ion complexes with cryptands 0211, 0221 and 0222 in water and several nonaqueous solvents. They used 7Li nmr technique and they found that the chemical shift of the lithium ion complexed by 0211 is essentially solvent and anion inde- pendent, indicating that the lithium ion is completely shielded by the cryptand molecule, as was expected. 0n the other hand, the chemical shifts of Li+—0221 and especially Li+-0222 complexes are solvent dependent. They also determined the formation constants of Li+-0222 in water and pyridine by using 7L1 nmr technique. The values obtained were log K = 0.99 i 0.15 for water and log K = 2.94 i 0.10 for pyridine. By using temperature variations in 7Li nmr measurements, they studied (26) the kinetics of complexation reactions of the lithium ion with cryptand 0221 in pyridine, water, dimethyl sulfoxide, dimethylformamide, and formamide and with 0221 in pyridine. Activation energies (Ea), rate constants (k1), and values of AHOail , Aso+ , and 00+ for the release of Li+ from the cryptates in the above solvents, are reported. Loyola £3 51. (27) used stopped-flow technique to measure the activation parameters for the formation of Ca2+ 1r- ‘1:— 9 complexes with 0222, 0221 and 0211, Sr2+ with 0222 and 0221 and Ba2+ with 0222 by using murexide and metal-phthalein indicators. They also determined the values for the dissocia- tion of the Ca2+ complexes of the three cryptands by using appropriate alkali metal ions as scavengers for dissociated cryptands. They conclude that the reaction order that they 2+ 2+ toward 0222 and Ca2+ 2 found Ca2+ < Sr < Ba < Sr + toward 0221 parallels the expected lability of the metal ion and indicates that desolvation of the metal ion is important before or at the rate—determining stage of cryptate formation. The pH dependence of formation of Ca2+, Sr2+ and Ba2+ complexes of 0222 has also been measured. Their results support the assumption of unreactivity of the protonated cryptands. Alkali metals can be solubilized in nonpolar solvents by the addition of an appropriate cryptand to the solution (28). Dye and coworkers (29,30) first found spectrophotometric evidence for the existence of alkali metal anions (Na', K") in amine and ether solutions in which they used cryptand or crown to dissolve alkali metals. They were able to crystallize the Na+-0222-Na' compound (31) by dissolving pure sodium in ethylamine (EA) and in THF in the presence of 0222. Due to the complexation of the cryptand with Na+, the concentration of dissolved metal was greatly enhanced and the gold-colored Na+-0222-Na' compound precipitated at low temperature and its crystal structure was determined (32). They were able to monitor the 23Na chemical shift of the sodium anion (33). At low temperatures they observed two nmr resonances, a broad 10 one for Na+-0222 and a narrow one for Na' which is about 63 ppm upfield from saturated aqueous NaCl solution. They extended the study and measured the 23Na nmr spectrum of Na+-0222-Na‘ as a function of temperature as well as that of 87Rb' in EA and THF and of 133Cs in THF. In addition to the study of the complexation of metal ions with macrobicyclic ligands (denoted as [21-cryptands), Lehn gt gl. also synthesized macrotricyclic ligands (denoted as[31-cryptands) (35-38). They are formed by two macrocycles linked by two bridges (Figure 0). They define three cavities: /'O\ /P “N N N' 4" \0/ “’ +‘2’ —-———b .0.— oul‘ ..__ O\“ MtQ‘o‘o /-°\ \ ”|"‘\\ VOK N O ,N N 6 ‘N Figure 0. Intramolecular cation exchange in mononuclear complexes of a macrotricyclic cryptand. two lateral circular cavities inside the macrocycles and a central cavity. The size of central and lateral cavities can be changed by modifying the size of the macrocycles and the length of bridges. They also form strong complexes with various metal cations. Their complexes are non-symmetric mononuclear and symmetric binuclear 3 -cryptates. Also 11 heteronuclear bimetallic [3l—cryptates can be obtained. The cation exchange rate between sites on two rings inside the cavity of a [31-cryptate has been studied by 130 nmr (39). The spectral changes observed agree with an internal cation exchange between "tOp" and "bottom" of the molecule as shown in Figure 0. Intermolecular exchange also occurs, but at much slower rate than the intramolecular one. Both intra- and intermolecular cation exchange is fast for the weak complexes of the ligand with alkali cations. Another [Bl-cryptand has been synthesized recently (00). As shown in Figure 5a, this molecule displays an attractive topology since it contains a Spherical intramolecular cavity Figure 5. Newly synthesized macrotricyclic cryptand. into which the substrate may be included. The Spherical cavity has ten binding sites in an octahedrotetrahedral arrangement (Figure 5b). The four nitrogens are located at the corners of a tetrahedron and the six oxygens are at the corners of an octahedron, whose centers coincide and their 12 ten corners lie on the same sphere. The logarithms of the stability constants for the K+, Rb+ and Cs+ complexes in water are 3.0, 0.2 and 3.0 respectively. The proton nmr kinetics study showed that the cation exchange rates are slow, and the activation energies for the dissociation of the complex are rather high, of the order of 16 kcal/mole. This ligand also forms very stable complexes with anions like 01' and Br— (00). The anion cryptates have been studied by 130 nmr. The structure proposed is shown below (Figure 6), and it has been confirmed by the determination of crystal structure (02). Figure 6. Anion cryptate. The ligand is in the tetraprotonated form and the anion is held inside the cavity. The logarithms of the stability constants of the chloride and bromide inclusion complexes 13 of the tetraprotonated ligand have been determined to be > 0.0 and < 1.0 respectively in water. They display remarkable Cl'/Br' selectivity. B. Study of the Thermodyggmic Parameters of the Complexation of Crowns and Cryptands As previously indicated, results of numerous studies of the stabilities of cation complexes of crowns and cryptands have been reported in the literature. From the formation constant, K, one can obtain the free energy of complexation AGC. Analysis of the free energy of complexation into enthalpy, and entropy, gives a better understanding of the thermodynamics of the complexation reaction. However, the enthalpy and entropy of alkali metal ion complexation have not been studied in much detail. Simon gt gt. (03-05) used a computerized microcalorimeter to study the thermodynamic properties of alkali complexes of various carrier antibiotics in methanol and ethanol. H. J. Moschler gt g;. (06) studied the thermo- dynamic parameters of the valinomycin—potassium iodide complex in ethanol by using microcalorimetric technique as well. R. Winkler (07) published a review article on the kinetics and thermodynamics of alkali ion complexes in solution. Frensdorff (08) in his original report of the thermo- dynamics of cyclic polyether complexation reactions, noted the remarkable increase in the stability of the complexes of cyclic polyethers over their linear counterparts. By comparing the complexes of Na+ and K+ with pentaglyme and l0 l8-crown—6 in methanol, he noted a 103 to 101+ enhancement of the stability constant in the cyclic ligand complexes. This phenomenon is described as the "macrocyclic effect" and the term was first postulated by Cabbiness and Margerum (09) who used this term in order to distinguish it from the chelate effect. Although it is well established that the chelate effect is of entropic origin, no agreement has been reached as to whether the macrocyclic effect is a result of a more favorable enthalpy or entropy terms in the cyclic ligand reactions. Cabbiness and Margerum (09) reported the macrocyclic effect for the cyclic tetraamines and proposed that ligand solvation and configuration were the important factors to be considered rather than the changes in trans- lational entropy. In a later publication, Hinz and Margerum (50) reported the detailed study of the thermodynamic properties of nickel(II)-tetramine complexes in water. They found that the enhanced stability of the cyclic as compared to the open— chain ligand (the macrocyclic effect), is almost entirely due to more favorable enthalpy changes. They attributed these changes to the decreased ligand solvation of the macro- cycle which has less amine-hydrogen—bonded water to be displaced in complex formation. Dei and Cori (51) offer the same explanation as above after studying the enthalpies of 2+ in water. Paoletti reaction of these same ligands with Cu gt gt. (52) conclude that the macrocyclic effect is due both to entropy and enthalpy terms, while Kodama and 15 Kimura (53) after studying the copper(II)-tetramine complexes by polarography state that the macrocyclic effect is due solely to the entropy term. The enthalpy and entropy changes for the complexation reaction of crown ethers have been studied and a discussion of the results will help clarify the nature of the macro- cyclic effect. There are distinct advantages in studying the macrocyclic effect using crown ethers as opposed to tetra— amine ligands. a) The ligands are uncharged at neutral pH and their complexing ability is not pH dependent, b) the reaction kinetics are rapid so that equilibrium measurements are readily obtained, c) among the metal ions which form complexes with these ligands are the alkali and alkaline earth cations which can be considered to be simple charged spheres, unlike transition metal ions which have specific stereochemical preferences. Izatt gt gt. (50,58) used a precision thermometric titration calorimeter to study the thermodynamics of formation of complexes of crown ethers. They studied the interaction of 15-crown—5, l8-crown-6 and of two isomers of dicyclohexo- 18-crown-6 with Na+, K+ Sr2+, Ba2+, Pb2+, ng , Rb+, 0s+. Ag+, T1+, NHL; CH3NH3+, 2+ ions in aqueous solutions at 25°C and u = 0.1. With the exception of Na+, Ag+ and Hg2+ + and Ca with the A isomer of dicyclohexo-lB-crown-6, all the reactions studied are exothermic (AH < 0). In many cases the entropy change of the complexation reaction is negative. Their data 16 contain no reproducible trends in AH or AS among the complexes of cations studied to explain the macrocyclic effect. They also performed (59) calorimetric titration studies of the 2+ 2+ and Pb2+ interaction between Na+, K+, Rb+, Cs+, Ca , Sr and the cyclic polyethers benzo—l5-crown-5, l8-crown-6, dibenzo-20-crown-8 and dibenzo-27-crown—9 in CHjOH-HZO solvents. In all cases negative AH and AS values for the complexation reaction were found. They conclude that as the cyclic polyether ring size increases, the AS values for the 1:1 reaction of a given cation become more negative, suggesting that significant conformation changes may be important in the formation of these complexes. Recently Izatt gt gt. (60) synthesized two compounds, 2,6-dioxo-18-crown—6 and 2,0-dioxo-l8-crown-6 similar to 18-crown-6 having carbonyl groups as valinomycin does and studied their thermodynamic properties with various cations in methanol. They found decreased stabilities of the new complexes due to less exothermic AH values, but more favor- able TAS values. By introducing a pyridine ring in l8-crown-6, they synthesized a new crown ether which forms unusually 2+ in methanol. stable complexes with Na+, K+, Ag+ and Ba They found that the entrepy term is the one favoring complexation while the enthalpy term works in opposition. Schori gt gt. (62) studied by conductance the thermo— dynamics of the complexation of various crown ethers with sodium salts in DMF and DME at different temperatures. They found that the complexation reaction in DME is both 1? entropy and enthalpy driven. E. Mei gt gt. (63.64) studied the kinetics and thermodynamics of the reaction of l8-crown-6 with 05+ in pyridine by using Cs—133 nmr technique. They found that the reaction occurs in two steps and that the enthalpy and entropy change for the second step are both negative. An extension of the macrocyclic effect is the cryptate effect which is described as the enhancement of the stability of the macrobicyclic complexes as compared to the macro- monocyclic ones. Kauffmann gt gt. (65) determined the enthal- pies and entropies of formation of alkali and alkaline—earth cryptate complexes from calorimetric measurements. Their results with the alkali cations are listed in Table 2. They show large negative enthalpies and sometimes negative 2+ 2+ entropies of complexation. The Sr and Ba as well as [Li+021l], [Na+0221]cryptates are of the enthalpy dominant type with also a favorable entropy change. The Ca2+ and (Li+0221] cryptates are entirely entropy stabilized with about zero heat of reaction. They conclude that the cryptate effect is of enthalpic origin. The enthalpies of complexation show selectivity peaks as the stabilities do, whereas the entropy changes do not. E. Mei gt gt. (66) studied the thermodynamics of the complexation of Cs+ with 0222, cryptand in acetone, propylene carbonate and N,N‘-dimethy1formamide by using Cs-l33 nmr technique. They found that two kinds of complexes exist in solutions. an inclusive and an exclusive one, and they calculated the enthalpy and entropy changes for both types. 18 .:.o mpwss hQOHPCm Ca m< oHos\Hmox ca z< QoM#Mthho mo moflmohpcm can mmamawanm .om< .om< impoz as- a: - o --- --- ma a 8.3 :.m- N.: - m - --- --- one Nmm --- m.mH- H.2H- a- --- oma Am w.mv --- m.HH- :.HH- 0.5- --- oma mmm uun m.o n m.: n m.w :.HH om< Am N.mv --- :.m . m.m . mm.nu o.o o=< Hmm --- --- --- m- m oma Ax c.HV --- --- --- :.m- H.m - oma Ham hmpmsmumm Am no.av+mo Ax me.Hv+pm Am mm.av+x Ax mo.ov+mz Am ms.ov+uq oaemcseoepwas seamen .0808: on comm pm seemeHooamo an Umuzmwws sowpommm Cowpmxmamaoo evapmhhoufiamxa< Mo mmfipapcmsa owamzhuoshosa .N canoe 19 Conclusion The transport of alkali ions by ionophors through artificial and biological membranes is a subject of consider- able interest to chemists and to biologists. It has been shown that cyclic polyethers (crowns) as well as naturally occurring antibiotic ionophores, such as valinomycin (67), are alkali ion transporters. Cryptates, where the alkali ion is completely enclosed into a hydrophobic sheath, should be at least as good ion transporters as crowns. However, diazapolyoxa cryptands are quite basic. For example, the acidity constants of cryptand 0222 are, Kl = 5.25 X 10'8 and K2 = 2.51 X 10'10 (18). Since the complexes can only be formed when the ligand is deprotonated, it is obvious that complexation reaction can only occur at high pH. As expected no ion transport was observed through black lipid membranes at neutral pH (68). Black lipid membranes are notoriously unstable at high pH's and attempts to measure ion transport in basic aqueous solutions were unsuccessful. The penultimate step in the synthesis of cryptand 0222 is the corresponding dilactam (Figure 2b). It is obvious that the compound is much less basic than the cryptand. Indeed, preliminary experiments seem to indicate that the 0222— dilactam (0222D) does transport alkali ions through black lipid membranes (68). It was of interest to us, therefore, to study the complexing ability and the thermodynamics of the complexation reaction of the 0222D in water and in nonaqueous solvents 20 in order to see how this change in the structure of the ligands affects its interactions with alkali metal ions. There are no reports in the literature on complexation studies of this ligand. '. CHAPTER II EXPERIMENTAL PART 0% Lithium perchlorate (Fisher) was dried at 190°C for several days. Sodium tetraphenylborate (Baker) and sodium perchlorate (G. F. Smith) were used without further purifica- tion except for drying. They were dried under vacuum at 60°C for 72 hours. Cesium thiocyanate (Rocky Mountain Research, Inc.) was recrystallized from absolute ethanol and vacuum dried. Cesium tetraphenylborate was prepared by a metathetical reaction between equimolar amounts of sodium tetraphenyl- borate and cesium chloride in a tetrahydrofuran-water mixture. The cesium tetraphenylborate precipitate was filtered and washed with conductance water until flame photometry registered sodium content of the order of conductance water. Drying was done under vacuum at 80°C for 08 hours. Anhydrous silver perchlorate (0. Frederick Smith Chemical Co.) was dried over P205 under vacuum and kept in the dark to prevent decomposition. Potassium hexafluorophosphate (Alfa Products) was recrystallized from water and dried under vacuum over P205 for at least 20 hours before use. The preparation of Li perchlorate is described in reference (69). B. Ligands The dilactam of cryptand 222 (0222D) was synthesized by a modification (70) of the method of Dietrich gt gt. (9) which has been described in detail (71). The ligands 2,2' bipyri- dine, 0,0'-dimethyl,2,2'-bipyridine, and 2,2'-biquinoline were obtained from 0. Frederick Smith and were dried under vacuum 0 21 22 C. Solvents Nitromethane (spectroscopic grade, Aldrich) was fractionally distilled over phosphorus pentoxide under reduced pressure and dried for 20 hours over freshly activated 5A Linde molecular sieves. Dimethyl sulfoxide (Fisher), was dried over Linde 0A molecular sieves for 2 days. Methanol (Fisher) was first fractionally distilled from calcium hydride and then from magnesium turnings in a nitrogen atmosphere. Dimethylformamide (Fisher) was vacuum distilled over P205. Propylene carbonate (Aldrich) was dried for 2 days over Linde 0A molecular sieves followed by vacuum distillation. Acetonitrile (Matheson Coleman and Bell) was refluxed over calcium hydride and then fractionally distilled over granulated barium oxide. Pyridine (Fisher) was refluxed over granulated barium oxide and then fractionally distilled in nitrogen atmosphere. Tetrahydrofuran (Baker) was dried over metallic sodium and benzophenone by refluxing. The molecular sieves used were activated by heating them at 500°C under dry argon for 12 hours. Analyses for water in salts and solvents, where possible, were carried out with an automatic Karl Fischer Aquatest II (Photovolt Corp.) titrator. In all solvents the water content was < 100 ppm. Important solvent properties and solvent abbreviations used in this thesis are listed in Table 3. ANNV cmvowumhm* 8.0 8.2 1%.? 8%? 03m: mmé- 0.0m wmé. $8.0. $1.5 28058822me mm.on H.mn 0:.NH NHw.on mcflvfihhm mH.on H.nH o.mw :mo.ou Aomv opmsonhmo osmahmonm My mo.ou m.m m.mm Hmm.on osmopmsonpwz no.0: *m.mm m.mm mam.ou Hocmnpms 85. 9mm 8.3 80.0- 325 83082522828 as- 8.8 Ram 2a.? 25 32.60320er mm.ou H.3H m.mm :mm.ou maflppfisopmo< AEQQV owl mpom>Hom .owuHom hmx .m canoe 20 D. Sample Preparation Since lithium salts are very hygroscopic, the water content of each solution was carefully maintained at the lowest possible level so that its total concentration remained less than 1% of the salt concentrations. All lithium, sodium and cesium salt solutions were prepared in a dry-box under nitrogen atmosphere. Dilute solutions of the salts were prepared by an appropriate dilution of a stock solution. Ligands were weighed out in the desired amount into 1 ml volumetric flask and then introduced into a dry-box for subsequent manipulation. E. Instrumental Measurements l. Lithium:7. Sodiug;23 and Cesium-133 NMR Sodium-23, lithium-7 and cesium-133 nmr measurements were made on a Fourier transform instrument using the magnet of a Varian DA-60 nmr spectrometer equipped with a wide-band probe capable of multinuclear operation (73), and computer controlled rf pulse generation and data collection which has been described previously (70). An external lH field look was used to maintain field stability. A Nicolet Instrument Corporation 1082 computer was used. The computer program (70) was used to generate a single rf pulse and to collect the resultant free induction decay (FID) signal. Data treatment was performed by the Nicolet FT-NMR Program (NIC-80/S-7202D) (75). The instrument was Operated at a field of 1.0092 T and at frequencies of 15.871 MHz, 23.318 MHz and 7.871 MHz for 23Na, 7Li and 133Cs respectively. 25 The references used were 0.0 M LiClOu in water and 3.0 3.0 M NaCl in water and 2.5 M NaClOu in methanol for the IS LiClOu in methanol for the lithium-7 measurements, sodium—23 measurements and 0.5 M CsBr in water and 0.2 M CsBr in methanol for the cesium-133 measurements. 10 mm nmr tubes were used. All the chemical shifts reported in this thesis are with respect to 0.0 M LiClOu in water and 3.0 M NaCl in water and infinite dilution chemical shift of the cesium ion in water. A positive value of 5 indicates a shift to higher field. The chemical shifts reported are corrected for differences in bulk diamagnetic susceptibility between sample and refer- ence according to the relationship of Live and Chan for non-superconducting spectrometers (76). _ g3 ref sample Scorr — debs + 3 (Xv ' Xv ) (l) where eref and Xvsample are the volume susceptibility of the reference and sample solutions respectively and éobs and 5 were calculated on the basis of published magnetic corr susceptibilities of various solvents (77). The magnitudes of corrections for various solvents are shown in Table 3. 2. Carbon—13 NMR Carbon-l3 nmr measurements were made on a Varian CFT20 Fourier transform nmr spectrometer equipped with computer controlled pulse generation and data collection. The 26 instrument was operated at a field strength of 1.8682 T and at a frequency of 20 MHz (78). For all 130 nmr studies, methanol was used as an external reference and D20 was used for locking the system. The solvent peak was used as a secondary reference. All chemical shifts reported are referenced to TMS. The sample solution was in an 8 mm nmr tube which was coaxially centered in the 10 mm nmr tube containing the external methanol reference with D20 for locking the system. 3. Far-Infrared Spectra The far-infrared spectra were obtained with a Digilab FTS-l6 spectrometer. The FTS-l6 is essentially a rapid-scan Michelson interferometer operated under computer control. The theory and operation of this instrument have been previously described (79). All spectra were obtained by using the 3- or 6- pm mylar beam splitters which cover the ranges of 600-150 and 025-100 cm'l, respectively. The spectra were obtained at a nominal resolution of 0 cm‘l, which gives a data point every 2 cm'l. The instrument was operated in the single beam mode. The reference spectrum was stored in the computer memory and subtracted from the solution spectra. Standard demountable cells (Barnes Engineering Co.) were used with 2-mm polyethylene discs, and the path length was maintained at 0.1 or 0.2 mm. All Spectra were smoothed by using the 9-point smoothing routine developed by P. R. Handy (79). 27 0. Data Handling Extensive use of the CDC-6500 computer was made to evaluate data. Program KINFIT (80) was employed to determine complexation constants. A linear least squares program was used to obtain enthalpies and entropies. CHAPTER III SPECTROSCOPIC STUDIES OF COMPLEXATION OF ALKALI METAL IONS, WITH THE DILACTAM OF 0222 A. Lithium-ZI Sodium-23 and Cesium-133 NMR Study of 0222D Complexes in Vagious Solvents Nuclear magnetic resonance (nmr) of alkali nuclei such as 7Li, 23Na, 39K and 133CS is one of the most powerful techniques for the elucidation of the nature of the various alkali species in solutions and the equilibria among these species. The chemical shifts and line widths of the nuclear resonances of alkali metal ions nuclei can give information about ion—ligand, ion-solvent and ion—ion interactions. Lithium-7 and cesium-133 nuclei are highly suitable for nuclear magnetic resonance studies because the resonance lines of Li+ and Cs+ ions in solutions are exceptionally narrow and chemical Shifts can be measured with considerable accuracy. Sodium-23 has a large quadrupole moment 2“ cmz), and its chemical shift range is rather (0.1 x 10' large (about 00 ppm). These two factors make 23Na nucleus a sensitive probe of the electronic environment around the nucleus as previous studies in this laboratory (81-80) and elsewhere (85-87) have Shown. 7Li, 23Na and 133Cs nmr have been found to be useful techniques for the determination of the formation constants of weak and medium strength complexes (25,88,89,90). The purpose of the study of this chapter is the investigation of the complexation reaction of Li+ ion, Na+ ion and Cs+ ion with the dilactam of 0222 in different solvents and the determination of the formation constants, where possible. 28 29 Alkali metal ion chemical shifts were determined as a function of dilactam/metal ion mole ratios and the results are shown in Tables 0-6. In all cases only one resonance of the metal ions was observed irrespective of the ligand/metal ion mole ratio. In general, if the rate of exchange of the metal ion between the two sites (free ion in the bulk solution and the complex), is larger than /2/nAv, (Av = difference between the characteristic resonance frequency in Hz in each site) only one population- average resonance is observed. The frequency of the lithium—7 resonance in water, dimethyl sulfoxide, dimethylformamide and methanol was found to be essentially independent of the ligand/Li+ mole ratio (Figure 7). It is evident that in these solvents the immediate environment of the lithium ion is not changed upon addition of an excess of ligand and, therefore, that at best only a very weak Li+-C222D complex is formed. It should be noted that the above solvents have a strong solvating ability as indicated by the respective Gutmann donor numbers of 33.0, 29.8, 26.6, 25.7 (91), and con- sequently in these media a relatively weak complexing agent is not capable of effectively desolvating lithium ion. The above results agree with previous observations that in dimethyl sulfoxide solutions (25) 0222 cryptand does not form a stable complex with lithium ion. 0n the other hand, as seen in Figure 8, in solvents of medium to weak donor ability such as acetone, tetrahydrofuran 30 Table 0. Lithium—7 Chemical Shifts at Various Mole Ratios, [0222D1/[Li+], in Nitromethane at 33 i 2° 0. Salt [Li+]M [0222D]/[Li+] 6(ppm) LiClOu 0.012 0.00 0.30 0.25 0.19 0.03 0.11 0.58 0.00 0.67 —0.12 0-75 '0-23 0,83 -0.30 0.95 —o.00 1.00 -0.52 1.21 -0.52 1.33 —o.52 1.50 -o.50 1.75 —0.50 2.00 —0.59 2051 -0060 31 Table 5. Sodium-23 Chemical Shifts at Various Mole Ratios, [0222D]/[Na+], in Various Solvents, at 25 i 2°C. Solvent Salt [Na+]M [0222D1/[Na+]6(ppm) Av%(Hz) Pyridine NaClOu 0.015 0.00 1.0 00 0.33 0.6 112 0.66 0.5 150 0.90 0.1 208 1.27 -0.1 217 1.60 0.1 270 2.11 0.1 272 CHBNO2 NaBPhu 0.05 0.00 10.5 16 0.09 7.3 212 1.00 0.7 607 1.27 5.1 605 1.99 0.8 606 CHBCN NaClOu 0.015 0.00 7.6 8 0-53 4-9 51 0.80 3.3 93 1.00 2.3 110 1.19 2.0 119 1.28 1.7 125 1.00 1.7 205 1.93 1.7 180 32 Table 5. Continued Solvent Salt [Na+]M [C222D]/[Na+] (S (ppm) Av%(HZ) MeOH NaClOu 0.015 0.00 0.0 18 0.06 3.6 20 0.93 3-7 39 1.00 3.5 68 2.20 3.0 120 H20 NaClOu 0.010 0.00 0.53 10 0.50 0.06 15 1.00 0.62 10 33 Table 6. Cesium—133 Chemical Shifts at Various Mole Ratios, [0222Dl/[Cs+], in Various Solvents, at 25 1 2°C. Solvent Salt [08+]M [0222D]/[Cs+] 6(ppm) DMF CsBPhu 0.015 0.00 0.30 0.00 - 0.28 1.13 0.30 2.00 — 0.28 CSSCN 0.010 0.00 50.62 0.27 50.59 0-59 46-33 0.99 02.69 1.30 39.82 1.60 36.87 2.13 33.07 2.70 30.12 3.20 28.57 5.20 22.21 020 CsBr 0.010 0.00 -1.07 0.96 -0.75 2.02 —0.00 CH30N CSSCN 0.010 0.00 -33.86 0.00 -32.93 0.77 -32.16 1.29 ~31.50 1.80 -30.92 30 Table 6. Continued Solvent Salt [05+]M [0222Dl/[Cs+] 0(PPm) CHBCN CSSCN 0.010 2.20 -3o.30 2.60 -29.98 2.99 -29.52 3-54 -29-37 0.25 -28.75 5.00 -28.90 pyridine CsBPhn 0.015 0.00 39.01 0.53 22.91 0.80 16.01 1.20 8.21 1.81 — 1.39 2.07 - 3.89 2.33 — 6.19 2.80 - 9.50 3.66 -12.69 0.37 -10.89 0.98 -16.09 10.00 —21.59 35 r DMSO I0 I I I MeOH A A A A a 0.0 U 3 H20 8 A A A = A EOI'NHZ (Ppm) / ' DMF ’ l-O ' /o/’_(fio co 2 o o/ _ J I 2'00.0 LO 2.0 —C2220il —Li+ Figure 7. Lithium-7 chemical shift as a function of 0222D/Li+ mole ratio in dimethyl sulfoxide, methanol, aqueous, formamide, dimethylformamide, and acetone solutions. (From Ref. 71). 36 3.0 THF N M PY - 3.0 l l 0.0 |.O 2.0 3.0 MCZZZDiI /MLi+ Figure 8. Lithium-7 chemical shifts as a function of C222D/Li+ mole ratio in acetonitrile,* tetrahydro- furan,* nitromethane and pyridine* solutions. (*From Ref. 71). 37 and nitromethane (resPective Gutmann donor numbers of 17.0, 20.0 and 2.7) lithium-7 chemical shift is affected by the addition of 0222D indicating formation of the lithium-0222D complex. Formamide (DN = 20.7) falls into the intermediate class. As seen in Figure 7, there is a small upfield shift with increasing mole ratio, indicating a weak interaction. The values of the formation constants for the 0222D complexes were calculated from the variation of the metal ion chemical shift with the ligand/M+ mole ratio. Assuming that only cation-ligand interactions are important, the observed chemical shift is a population average of the two chemical environments of the metal ion as shown in eq. (2), dobs = GMX‘M + GMLXML (2) where Gobs is the observed chemical shift, XM and XML are the fractions of the free and complexed metal ion respectively while 5M and 6ML are the respective chemical shifts for the two species. Assuming a 1:1 complex, we have the equilibrium + M+L+ML (3) where L is the ligand. The formation constant of the complex, in concentration units, becomes K = CML CMCL (0) 38 where CM and CML are the equilibrium concentrations of the free ligand and the complex respectively. 2 2 _ t t 2 t 2 t obs—(KCM-KCL-l):(KCL +KCM - 2 t t t t a; 2K 0M0L + 2K0L + 2K0M + 1) - 0 5M L ——-—-+ (5) 2K0t 6ML M In eq. (5), 0; and 0;, the total concentration of the metal ion and of the ligand respectively, are known and 6M can be easily determined from measurements on solutions of lithium salts without the ligand. Eq. (5) then contains two unknowns K and 6ML' In the case of a rather strong complex §WL can be determined experimentally by the addition of such excess of L that essentially all of the metal is complexed. In the case of weak complexes eq. (5) is solved by the following procedure. The experimental parameters Gobs’ 0;, CE and 5M are substituted into the equation, and K and 5ML are varied until the calculated chemical Shifts correspond to the experimental values within the error limits. The data were analyzed on a CDC-6500 computer using the FORTRAN IV program KINFIT (80). Weighed input data were used for concentration and chemical shifts. The results of the above calculations are shown in Table 7. 0 For more stable complexes with Kf > 10 the chemical shift-mole ratio plot consists of two straight lines 39 intersecting at 1:1 mole ratio. Such a plot cannot be analyzed by our equation, and in such cases we can only conclude that Kf > 100. This behavior was observed for Li+-dilactam solutions in nitromethane which is not surprising since nitromethane is a very poorly solvating solvent with Gutmann donor number of 2.7. A more unexpected behavior was observed in pyridine solutions where the lithium—dilactam complex formation con- stant was found to have a respectable value of log Kf = 2.60. Yet pyridine should be a good solvating solvent as indicated by the donicity of 33 and the magnitude of its sodium-23 chemical shift (92). It is possible however, that pyridine, being a nitrogen donor, or a "soft base". does not solvate strongly a "hard acid" such as Na+ ion (92). A recent pmr study of Na+ ion solvations in nonaqueous solvents by Ahmad and Day (93) strongly supports this conclusion. The lithium-dilactam complex is only slightly more stable in acetonitrile and tetrahydrofuran solutions than in the solvents mentioned above. In the former case replacement of the perchlorate counterion by bromide did not influence the formation constant of the complex. As seen from Figure 8, 7Li chemical shifts seem to converge at high dilactam/Li+ mole ratios. These results seem to indicate that the lithium ion must be inside the dilactam cavity but that it is not insulated completely from the solvents as was the case of Li+ ion enclosed in the 0211 cryptand (25). 00 The study of the sodium complex with 02220 by 23Na nmr was complicated by the limited solubility of the complex in most nonaqueous solvents and also by the quadrupolar broadening of the 23Na resonance due to the unsymmetric structure of the dilactam. Some of the linewidth at half height are «.700 Hz and, therefore, the measurements of the chemical shift are much less precise than in the case of 7Li. It should be noted that the quadrupole moment of 23Na is considerably higher than that of 7Li (0.1 X 10'2“ cm2 compared to -0.2 X 10"26 cmz). Solvents like pr0pylene carbonate (PC), acetone, diglyme, benzene, chloroform, tetrahydrofuran and ethylenediamine were tried, but the solubility was very limited. It is natural to eXpect that Na+ ion will form stronger complexes with 0222D than Li+ since its size is closer to the dimension of the dilactam cavity. Indeed, as shown in Figure 9, sodium-23 chemical shifts-~dilactam/Na+ mole ratio plots Show in methanol, dimethylformamide, nitromethane, aceto- nitrile, and pyridine solutions sharp breaks at 1:1 mole ratio indicating that Kf > 104. The limiting chemical shifts of the complexed sodium ion are solvent-dependent indicating incomplete insulation of the cation from the solvent. In dimethyl sulfoxide solutions the 23Na resonance line becomes extremely broad upon addition of 0222D. Consequently, the chemical shifts cannot be measured accurately and no definite conclusions can be made on the stability of the dilactamrNaf complex in this solvent. 01 l5.0 I0.0- 8 (ppm): ‘ l . ‘\ .0 M?OH ‘s‘w- ‘ .- DMF ‘I '. . _ ACN I42C> 0.0 + H” , J I 0.0 IO 20 M02220 / MNd‘ Figure 9. Sodium-23 chemical shifts as a function of 0222D/Na+ mole ratio in nitromethane, methanol, dimethyl- formamide, acetonitrile, aqueous and pyridine solutions. 02 The large cesium ion does not fit conveniently into the 0222D cavity. As indicated in Figure 10, no 133Cs chemical shift is observed upon addition of the dilactam to cesium salt in DMF solutions. In the case of nitromethane and pyridine solutions (Figure 10) and acetonitrile solutions (Figure 11), addition of the dilactam produced large but gradual downfield shifts indicating formation of weak complexes. The large difference in the limiting chemical shifts of the Cs+-0222D complex in the above solvents is a good indication that the metal ion remains eXposed to the solvent. Computer analysis of the data for the cesium—dilactam system gave log K = 1.67 t 0.02 in nitromethane, log K = 1.96 t 0.01 in pyridine and log K = 1.68 t 0.08 in CH CN. 3 It should be noted, however, that in all three solvents cesium salts are ion paired. Therefore, the formation constants given above reflect competition between the anions and the dilactam for the cesium ion. In pyridine solutions it was found that the ion pair formation constant for Cs+BPh4' is 370 :q20 (90) while for CSSCN in nitromethane and acetonitrile it is 00.1 t 1.2 and 13.2 i 3.8 respectively (95). Comparison of the above values of the formation constants of the cesium-0222D complexes with the ion pair formation constants makes it quite apparent that the anion should be taken into consideration as both ligand and anion compete for the cesium ion in solution. The observed chemical shift 43 60.0 50.0 >.\ _ 4CM3" ‘§\‘( \A ‘\\“\\\~ 3 . - 8 00 A\A 20.0— \ . O H¢3£)—' \\\. [MAP 0.0gl—3 \Hf" 'I0.0"' \.\ o \.\0 FY 2 o I I I I I O 0.0 LO 2.0 3.0 4.0 5.0 Mczzzon / More Figure 10. Cesium-133 chemical shifts as a function 'bf C222D/CS+ mole ratio in nitromethane, dimethylformamide, aqueous and pyridine solutions. 00 -300 (pm) -310 I 1 10 4.0 M02220 bécf' Figure 11. Cesium-133 chemical shifts as a function of 0222D/Cs+ mole ratio in acetonitrile. 45 therefore, is modified to include a third factor, the chemical shift characteristic of the ion pair: dobs = XM‘SM + XMLGML + XMA‘SMA (6) The derivation of the final expression for the observed chemical shift can be obtained by applying the following equations Kip = [MAI/[M] [A] (7) Kf = [ML]/[M][L] (8) 0L = [ML] + [L] (9) CM = [ML] + [M] + [MA] (10) [A] = [ML] + [M] (11) where Kip is the ion pair association constant, Kf is the formation constant of the complex, [MA], [M] and [A] are the concentrations of the ion pair, free metal ion and free anion respectively, and CL and CM are the analytical con- centrations of the ligand and salt. With these five equations and simple algebraic manipulations, the following expression is readily derived. 3 2 KfKip[M] + (KfKipCL + Kip + Kf)[M] + (KfCL + l - KfCM)[M] - 0M = 0 (12) 06 A detailed derivation of the above equation and the descrip- tion of the subroutine EQN used in KINFIT are given in Appendix II. In order to use this model, the ion pair formation constant has to be known. Using the values reported above, new formation constants were obtained for the cesium-0222D complexes and the values are reported in Table 7. Comparing the new formation constants obtained considering the ion pairing with the previously obtained ones we see that they are larger especially in the case of pyridine where the difference is about an order of magnitude. It is quite evident that for the weak cesium-0222D complex, 'the anion and the solvent can have strong influence on the complexation interaction. The 13303 chemical shifts of the complex in the above solvents do not approach the same value, which seems to indicate that the cesium ion is exposed to the solvent. If we compare the cesium-0222D complexes to the cesium-0222 ones (90), we see that the former are much weaker. It seems that the two carbonyl groups in the 0222D molecule make it more rigid than the 0222 molecule and thus it is more difficult for the large cesium ion to approach the binding sites of the ligand. It should be noted that in aqueous solutions all three cations studied above do not Show evidence of complex formation. This apparent contradiction of the results obtained in the black lipid membrane transport eXperiments can perhaps be explained by the fact that very small amounts of the complex can significantly lower the resistance of the 47 Table 7. Formation Constants of 0222-Di1actam Complexes in Nonaqueous Solvents at 25°C. Salt Solvent log Kf LiClOu Pyridine 2.60 : 0.10a LiClOu Tetrahydrofuran 3.12 1 0.09a Li0104 Acetonitrile 3.23 : 0.07a LiBr Acetonitrile 3.13 1 0.12a Li010,L Nitromethane > 0a NaPhuB Dimethylformamide > 0a NaClOu Pyridine > 0b NaClOu Acetonitrile > 0b NaPhuB Nitromethane > 0b CSSCN Nitromethane 1.67 : 0.02b CSSCN Nitromethane 1.79,: 0.03c CsPhuB Pyridine 1.96 1 0.01b CsPhuB Pyridine 2.86 i 0.010 CSSCN Acetonitrile 1.68 t 0.08 CSSCN Acetonitrile 1.70 10.080 8330C b25°C 0Corrected for ion pairing 08 6 black lipid membranes (BLM). For example, 10- M valinomycin found to change the BLM resistance from 109 ohms to m 105 ohms (68). Our alkali metal nmr data do not permit the differentiation between inclusive and exclusive types of cryptate complexes. An inclusive cryptate complex refers to one in which the metal ion is totally within the cavity of the ligand. In all cases, however, there is some influence of the solvent on the limiting chemical shift of the complexed cation, the influence being in the order Li+ < Na+ < 05+. In view of the 13303 results and the size considerations, it seems plausible to assume that in the last case the complex may be of the exclusive type. All the above values are the concentration constants. However, since the complexation reaction M+ + 02220 I M+-0222D (13) does not involve separation of charges, these values should represent reasonable approximations of the thermodynamic constants. B. pH Dependence o: the 23Mg30hemiggtt§gttt It was mentioned above that in aqueous solutions 0222D does not Show evidence of complex formation with lithium, sodium or cesium metal ions. The metal ion chemical shift is independent of the ligand/M3 mole ratio. It was interesting 09 to find out if changing the pH of the solution caused any change in the interaction of 0222D with the metal ion. A 0.01 M NaCl and 0.01 M 0222D aqueous solution was prepared and the measured pH of the solution was 7.85. The sodium-23 chemical shift of the solution is 0.15 ppm. By adding 0.096 M (Bu)4N0H into the solution of the pH of the solution was changed and the 23Na chemical shift was monitored at different pH values with the results shown in Table 8. Altering the pH of the solution within the range of 11.50 and 1.25 the 23Na chemical shift remains constant, which indicates that there is no change in the interaction between the ligand and the sodium ion. The pH of the solution was changed by the addition of 6 M_HCl. The same experiment was performed by using cryptand 0222 as the ligand with the results listed in Table 8. At the mole ratio of [0222 ]/ [Na+] = 1.0 essentially all Na+ is complexed and the 23Na resonance line has a chemical shift of 8.0 ppm and a linewidth at half-height of approximately 70 Hz. At mole ratios less than one, the 23Na resonance line is an unresolved doublet. To study the pH dependence of the 23Na chemical shift, a solution of 0.05 M NaCl and 0.05 M 0222 in water was prepared. The pH of this solution was measured to be 10.68. Adding H01 to the solution to lower the pH, the Na+-0222 peak broadens and at a pH value of 7.70 starts to split. As the pH is lowered further, the peak characteristic for the free Na+ grows larger and that of the complexed becomes smaller until it completely 50 Table 8. pH Dependence of 23Na Chemical Shift for Aqueous Solutions of Sodium Chloride. Ligand [Na+] [Ligand]/[Na+] pH A5(ppm) 10%(02) 02220 0.01 1.0 7.85 0.15 15 9.15 0.15 15 9-55 0-15 15 10.11 0.15 15 10.50 0.31 15 10.90 0.30 15 11.50 0.31 15 1.25 0.31 15 0222 0.05 1.0 10.68 8.0 70 8.55 6.8 95 8.22 6.8 108 8.00 6.3 120 7.70 3.5 7-55 1-23 7.00 0.3 27 6.35 0.0 15 1.65 0.0 15 51 pH =10.“ pH=170 pH=700 pfi=755 pH=605 10.0 51.0 0.0 p p m 5.0 090 Figure 12. Changes in Na+ resonance as the relative amounts of free and complexed Na+ vary with changing pH for 0.05 M aqueous NaCl with 0222 / Na+ = 1.0. 52 disappears at a pH of 6.35. Figure 12 shows the changes in chemical shifts as more of the Na+ is released from the Na+- 0222 complex on changing the pH of the solution. At this pH value according to the results of Mei-Tak Lok (96) all cryptand is in the diprotonated form. Lowering the pH further produces no observable change of the Na+ peak. Apparently the di- protonated crypt does not complex sodium at all. For the monoprotonated crypt no conclusion can be drawn, because at the pH level at which it exists, all three Species CNa+, CH+, CH2++ exist in solution and exchange among them is rapid. From the above experiments it is evident that the acid-base properties of 0222D and 0222 are quite different. C. pKa Determination of 0222D To determine the pKa of 02220. 25 m1 of a 0 x 10‘3 M solution of 0222D in conductance water were titrated with 5 X 10'2 M H01. Tetramethyl ammonium bromide was used to maintain constant ionic strength at 7 X 10'2 M. Nitrogen was bubbled through the solution to eXpel the carbon dioxide present. The results obtained are plotted in Figure 13. As is seen from the graph, there is a sudden dr0p of the pH of the solution from 7.28 to 0.16 after the first drop of H01 was added and after that the pH changed smoothly. It was expected that the titration curve would Show two breaks, one after the addition of 2.0 ml H01 and another after the 53 .Hom_a no.0 seas QNNNU moo05dm s m-oa x s so as o.m~ so compasses one you oesso soapmspae ma am . . . as as q. no: .2 . on . so .. - . . . ..w - .- - ea 1 oi 1 In so 50 addition of 0.0 ml H01. There is no indication of the breaks in the titration curve of Figure 13 which means there is no indication of protonation either of the nitrogens or of the carbonyl oxygens of the amide. 25 m1 of 7 x 10'2 M (Me)uNBr were also titrated with 5 X 10'2 M H01 solution. The shape of the titration curve is exactly the same as the one for the titration of the 0222D. The only difference is that the initial pH value of the blank solution was 5.0. The above results Show that the acid-base properties of the 0222D are quite different of those of 0222. This was expected to a certain degree since 0222 is an amine and, therefore, it is highly basic while 0222D is a diamide which is almost neutral. The above behavior of the C222D, i.e. no indication of protonation, supports the results obtained by 23Na nmr. Since the 0222D does not protonate on lowering the pH of the solution a change in pH should not affect the interaction of the 0222D and sodium ion. D. Far Infrared Study of Lithium_gnd Sogtum_00mplexes 0110 0222D in Nonaqueous Sgtvents Far infrared SpectroscOpy has been used extensively (69,97-107) for the investigations of the motion of alkali metal cations relative to their immediate environment. In general, variations of the cation's motional frequencies in solution are expected to occur with variations in the immediate environment of the cation, and are indicative of interactions occurring in solutions. 55 When a macrocyclic polyether complexes an alkali metal ion, it insulates it from the medium and a cation-ligand vibration is observed. Tsatsas gt gt. (108) investigated the far infrared spectra of the sodium and potassium-dibenzo- 18-crown-6 systems in several solvents and found a band whose frequency was solvent independent. The authors attributed this band to the vibration of the metal-ligand bonds. Cahen and Papov (109) studied the far infrared Spectra of sodium- 0222, sodium-0221, lithium-0211 and lithium-0221 cryptates 1 for the Na+-0222 vibration, l and found bands at 230 t 2 cm- at 218 1,1 cm'1 for the Na+-0221 vibration, at 308 t 1 cm- for the Li+-C211 vibration and at 203 i 3 om‘l for the Li+-0221 vibration which could not be assigned to the free metal or ligand. All the above bands were found to be solvent and anion independent, and were assigned to the vibration of the metal ion in the cryptand cage. In the present study far infrared measurements were carried out on the Li+-0222D complex in acetonitrile and pyridine solutions and on Na+-0222D in dimethylformamide and dimethyl sulfoxide. These are the only solvents in which the solubilities of the complexes were high enough to allow far infrared meaSurements. The spectra of LiClOu solutions in the two solvents show the band due to the vibration of the cation in the solvent cage at 005 cm'1 in acetonitrile and at 385 cm"1 in pyridine. Upon addition of one equivalent of dilactam to these solutions, the above bands disappear and a new band appears at 56 1 N 060 cm' whose frequency is independent of the solvent. 1 It is believed that the 060 cm' band is indicative of the vibration of the cation in the cavity of the ligand. IsotOpic substitution of 7Li by 6Li shifted this band to higher frequency. The spectra of NaBPhu in DMF and DMSO Show the band due 1 to the vibration of Na+ in the solvent cage at 190 cm' and 1 at 205 cm' respectively. Upon addition of the dilactam to 1 Na+ salvation band disappeared while a new band appeared at 125 cm'l. This the NaBPhu in DMF solution, the 190 cm- latter band represents the vibration of the Na+ in the 0222D cavity although it is difficult to understand why the band shifts to lower frequency. In the case of Na+-0222 cryptate the vibration of the Na+ cation in the cryptand cavity was observed at 230 cm”1 (109). In the case of DMSO the spectrum does not Show any change upon addition of the dilactam. The 1 solvated Na+ band remains unchanged which indicates 205 cm- that there is no change in the immediate environment of the sodium ion. This result agrees with the 23Na nmr data (71) that the 23Na chemical shift in DMSO was independent of the ligand/M+ mole ratio. Unfortunately the far infrared study cannot be extended to other solvents because of solubility problems. From the above data it cannot be concluded if the Li+-0222D and the Na+-0222D complexes are definitely of the inclusion type and if the solvent influences the frequency of the vibration of + . . . . the M -0222D. There is only an indication that the metal ion is inside the ligand cavity. 57 E. Carbon-l3_NMR Study of Lithium, Sodium and Potassium Complexes with 0222D in Nitromethane The carbon-l3 spectra of the free ligand 0222D and of the complexes Li+-0222D, Na+-0222D and K+-0222D in nitro- methane were obtained. The free ligand has nine different types of carbon atoms, which are indicated in Figure 10. Therefore, nine resonance lines are eXpected, one for the carbonyl carbon atoms, six for the 0-0 and two for the N-C carbon atoms. Since 0222D is an amide the nitrogen atom and the carbonyl group are c0p1anar and one of the two strands attached to the nitrogen atom is gtg and one is tpgpg to this plane. The exchange between them is slow and this makes the carbon atoms of the two strands nonequivalent. The carbon-l3 Spectrum obtained by using the CFT-20 nmr spectrometer shows seven resonance lines for the free ligand (Figure 10a). The carbonyl peak is not shown in the figure. The peak at 63.31 ppm is the nitromethane peak and the 50.97 ppm peak is the methanol peak used as the external reference. The four peaks downfield from nitromethane are assigned to the 0—0 and the two peaks upfield from nitro- methane to the N-C carbon atoms. Since six resonance lines for the 0-0 carbon atoms were expected but only four were obtained, the carbon-l3 spectrum of the 0222D in nitromethane was taken by using the Bruker-180 nmr Spectrometer which has higher resolution than the CFT-20. The new spectrum indeed shows nine resonance lines (Figure 15). The 72.03 ppm and 58 ‘ N‘“\¢0“\\\;f\\VN 0'20 MUG MOO" I 9 I b n 1 L 70 60 50 ppm Figure 10. Carbon-l3 spectra of: a) 0.10 M 0222D in CH3N02; N0. b) 0.10 M 02220 with 0.2 M LiClOu in 003 2 59 .Amoaosoppooam m 00x30 m m :4 0 son comm Iconsm H woman 22 Ome mv oz 30 .. QNNNO .2 mN.O .H n... NH O m . m on Oh . EQQ . Oh— J%. L . . Ayfi. 60 the 70.20 ppm peaks are both doublets but the chemical Shift difference is only 0.17 ppm in the former and 0.20 ppm in the latter case and the resolution of our instrument is not high enough to separate them. The carbon-13 nmr spectrum of the 2:1 [Li+]/[0222D] complex was obtained (Figure 10b) by using the CFT20 NMR spectrometer. At this mole ratio all ligand is complexed. The Li+-0222D spectrum also shows six 0.0 peaks all of them upfield from the ones of the free ligand, and two N-C peaks downfield from the ones of the free ligand. The lithium ion should be inside the ligand cavity but it is small enough not to change the conformation and the symmetry of the ligand. The Na+-0222D spectrum (Figure 16a) and the K+-0222D spectrum (Figure 16b) Show only two peaks in the 0-0 region and one at the N-C region. Sodium and potassium ions are much larger than the lithium ion and as they enter the ligand cavity they make the molecule more rigid. As a result of this the two strands attached to the nitrogen atoms cannot be gtg and tpgpg to the nitrogen-carbonyl plane anymore but have an intermediate position. At this position they can exchange rapidly between the two positions making many carbon atoms equivalent. Another assumption is that since the molecule is more rigid after complexation with sodium or potassium ion the conjugation between nitrogen and carbonyl is destroyed and the double bond is on the carbonyl group only. In such a case, the infrared absorption band 61 MeOH MW) WIN-W MeOH MWWMII 00011.4(. WWW!“ J *5 WW Figure 16. Carbon-l3 spectra of: a) 0.05 M 0222D with 0.06 M Pkfl3¢u 1J1I2H3 KPF6 in CHBNO 2. N02; b) 0.05 M 02220 with 0.1 M 62 IIOO tooo"” 1800 1000 Figure 16A. IR Spectra of the Carbonyl Absorption of: a) 0222D in nujol; b) Na+-0222D in nujol; c) [Li+I/[0222D] = 2.0 in CH N02; and 3 d) [K+]/[0222D] = 1.6 in CH N0 . 3 2 63 l for the free for the carbonyl group which appears at 1600 cm- ligand, should move to higher frequency. The infrared spectra of Li+-C222D, Na+-0222D and K+-0222D in nitromethane do not Show any new peak at frequencies higher than than 1600 cm"1 (Figure 16A). This proves that our first assumption was correct. From the carbon-l3 study it seems reasonable to conclude that all three cations studied above form inclusion type complexes with the 0222D in nitromethane. CHAPTER IV STUDY OF THE THERMODYNAMICS OF THE COMPLEXATION REACTION OF C222D WITH LITHIUM AND CESIUM IONS IN VARIOUS SOLVENTS INTRODUCTION A deeper understanding of the thermodynamics of the complexation reaction may be provided by dividing the free energy of complexation, AC, into enthalpy, AH, and entropy, AS, of complexation. There are altogether four possible combinations of the thermodynamic quantities leading to stable complexes (AC < O): a) AH < 0 and dominant, 038 > 0; b) AH < O and dominant, TAS < 0; c) TAS > 0 and dominant, AH < 0; d) TAS > 0 and dominant, AH > 0 (65). For electrostatic complexes between charged ligands and "hard" cations, TAS is generally positive and dominant. These are entrOpy stabilized (110,111). For complexes with more covalent character like the ones of the "soft" cations, AH is generally negative and dominant (110,111). In the case of the alkali metal cryptates enthalpic type behavior is eXpected because the ligands are uncharged. However, entropy dominant behavior is also possible since alkali metal cations are "hard" cations. The strength of the complexes of cryptand 0222D with alkali metal cations was discussed in Chapter III, as well as the solvent and the size of the metal ion effect on the complexation reaction. In this thesis an evaluation of thermodynamic quantities of the complexation reaction is attempted. To obtain the needed information, the effects of the variations of mole ratios and temperature on the 13303 and 7L1 chemical shifts of the Cs+-C222D and 60 65 Li+-0222D complexation reactions in pyridine, nitromethane and acetonitrile were determined. RESULTS AND DISCUSSION The complexation of cesium and lithium ion with 0222D cryptand can be described by the following equilibrium, assuming that a 1:1 complex is formed, M+S + LS I M+L S + ( + )S x y z x y - z where S is the solvent molecule, L is the ligand and x, y and z are the solvation numbers of the cation, ligand and the complex respectively. The variation of chemical shift with changing temperature was studied for the above complex- ation reaction. The systems studied were LiClOu and 0222D in acetonitrile, nitromethane and pyridine, CsBPhu with 0222D in pyridine, CSSCN with 0222D in nitromethane and CSSCN with 0222D in acetonitrile. At each temperature the chemical Shift of the metal ion was monitored as a function of dilactam/metal ion mole ratios. In all cases studied, only one p0pu1ation average signal was observed which indicated that the exchange of the metal ion between the two sites, that of the free ion in the bulk solution and that of the complexed ion, was faster than 107 per second. All chemical shifts reported in this study have been corrected for bulk magnetic susceptibility of the solvent and temperature effects on the reference solution. The 7L1 chemical shifts have been adjusted to the final reference 66 point of the 0.0 M aqueous LiClOu at 25°C, while the 1330s chemical shifts have been adjusted to the final reference point of the infinitely dilute concentration of the cesium ion in water at 25°C. The linewidths observed for the uncomplexed cesium and lithium ion are narrow (about 3-5 Hz) and are limited by the inhomogeneity of the magnetic field. No line broadening due to complexation is observed, the only exception being Cs+-0222D in pyridine where at low temperatures some broadening is observed, the maximum line width being on the order of 50 Hz. All the lithium-7 chemical shifts measured at various temperatures are listed in Tables 9-11 and shown in Figures 17-19. The formation constants were computed by the KINFIT program and the results are listed in Table 12. In all three solvents studied, i.e. acetonitrile, nitromethane and pyridine, the curvature in the mole ratio plots increases with increasing temperature, indicating an endothermic behavior. The calculated formation constants show the same trend (Table 12). In acetonitrile log K = 3.19 i 0.07 at —05°C and at 28°C it is 3.50 i 0.25 while at 86°C the complex is even stronger and from data simulation we get only an estimate of its value about 3.90. In nitromethane the calculated formation constant is log K = 2.93 t 0.06 at -29°C. At temperatures 3 27°C the mole ratio plots consist of two straight lines which intersect at the 1:1 mole ratio. In this case the formation constants cannot be calculated by the equation normally used and 67 00.0- no.0 00.0 00.0 00.0 H0.H 0m.o No.0 m0.H no.0- no.0 mo.o mm.o 00.0 H0.H 00.0 00.0 H0.H no.0- 0m.o mn.o no.0 00.0 0H.H 00.0 00.0 HN.H no.0 ms.o No.0 NH.H 0H.H 00.0 mm.o no.0 00.0 0m.o 00.0 00.0 0m.e HN.H 00.0 00.0 no.0 00.0 00.0 00.0 HN.H so.a 0m.a ss.o 00.H 00.H ss.o no.0 sm.H om.a os.a so.a 00.0 00.H sm.a No.0 0m.H 0a.a 00.H 0H.m oa.m 00.0 ms.a oo.m mm.o 00.0 Hm.m mm.m ms.m 00.m om.o oo.m mo.m sm.o ms.m oo.m mm.m 0m.m oo.m 00.0 mm.m Ho.n 00.0 000 00m om cam- omo- on- 000- capsm oo .opzpmhomsoa oaoz 0o .oHSPmnonoe macs .moMSPMMoQEoB mzowpm> as 20000 on 00000 no oesooosa one as as 000.00 000000 00 000000 Hooasoso 0.00 .0 00000 68 00.0- 00.0 00.0 00.0 00.0 00.0 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 000 000 00 000- 000- 00000 00. 000- 00000 00 .muzpmpmmEma 0002 Do .mMSPmquEwe @002 cmscfipCOU .0 00909 69 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0 -- -- 00.0 00.0- 00.0- 00.0- 00.0 00.0 00.0 -- 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 -- 00.0 00.0 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0- 00.0- 00.0 00.0 00.0 00.0 00- 000- 000- 00000 0000 000 000 00 000- 00000 00 .00300009809 0002 00 .mpzpmpmmsme macs N oz n .mmuSPmmmmEme m30000> #0 mo :0 QNmmo mo 00:00000 0:0 :0 08 000.00 300000 mo 0000sm 0000Emno 0-00 .00 00909 O .- «UT—A- U H,-H\A.LA ... H gunman-no .0 O .3 "30¢ 0.03-Lg 3:0. ...-4 A 2 Av NW . by V dwfiv an-.m..H .HAU «:00..h fl :mu Huwfuflntww 2”» RI Hg 0 ..H N mN-AN “.5 7O mo.ml mm.H| ou.on H0.OI ma.o oo.m 00.0- 00.0- 00.0- 00.0- 00.0 00.0 00.0- 00.0- 00.0- 00.0- 00.0 00.0 m©.N- mm.al 00.0- H0.0| ma.o wo.N mo.ml mm.H| on.on H0.0I mo.o om.a m©.NI mm.H- 00.0- H©.o- no.0 mm.H m©.N- mm.H| 00.0- Hw.ou mo.on 0N.H om.N- NO.N- om.o- 05.01 mH.OI 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0 00.0- 00.0- 00.0- 00.0- 00.0- 00.0 mm.m- mm.m- 0o.m- mw.0- mm.0- 00.0 000 000 0mm 00 0mm- 000mm 00 .00300009509 macs .000500000609 mso0nm> 00 00000000 00 00000 00 00000000 000 :0 02 000.00 000000 00 000000 00000000 0-00 .00 00000 71 ‘00 . ACN (:0 9?") O. O 1.0 2.0 3.0 Figure 17. Lithium-7 chemical shifts as a function of CZZZD/Li+ mole ratio in acetonitrile at various temperatures. 21) 72 ‘\ 3,7300% 0.0 . '— 00cc +427 C 1 + 80°C —2.o —- 0, 0 +102°c L l 4 (1C) 13) 2J3 313 1!. C222 0 /y_._ '+ Figure 18. Lithium-7 chemical shifts as a function of C222D/Li+ mole ratio in nitromethane at various temperatures. 73 PY’ . + 90°C l J 21) 31) +. Figure 19. Lithium-7 chemical shifts as a function of C222D/Li+ mole ratio in pyridine at various temperatures. 74 it can only be concluded that log Kf 3 4. In the case of pyridine the same endothermic behavior is observed. Unfortunately, no values for the formation constants were obtained in this case because the two unknowns in our equation seemed strongly coupled and it was not possible to calculate the formation constants at different temperatures. From the curvature of the mole ratio plots we can only con- clude that formation constants become larger as the temperature increases. The relevant thermodynamic parameters are calculated by the following relationships: AG = -RT an (14) AG = AH - TAS (l5) _A_§ Alli an - R - RT (16) By plotting an vs l/T a straight line is obtained with a AS slope of —A% and an intercept of +§— provided AH is indepen- dent of temperature. Such a plot is shown in Figure 20 for the Li+—0222D complex in acetonitrile and nitromethane. The lines obtained are straight. By linear least squares fitting of the data the lepe and intercept of the lines are calculated and the correSponding thermodynamic quantities are listed in Table 17. In acetonitrile AH = 1.58 i 0.22 kcal/mole while AS = 21.7 i 0.87 while in nitromethane AH = 6.10 i 0.16 and AS = 38.4 i 0.62 cal/mole-OK. In both cases the complex is enthalpy destabilized but entrOpy stabilized. A possible eXplanation is that Li+ is a small 75 Table 12. Formation Constants for the Complexation of LiClOu by C222D in CHBNO2 and CHBCN at Various Temperatures. TOC TOK 103/T log K In K System: LiClOu + CZZZD in CHBNOZ —29 244 4.10 2.93: 0.06 6.74 -20 253 3.95 3.14:;0.07 7.23 -13 260 3-85 3-27 : 0.09 7-53 - 7 266 3-76 3-38 :_0.06 7-78 0 273 3.66 3.85 i 0.33 8.86 27 300 3-33 >8 --- 80 353 2.83 >4 ——- 102 375 2.67 >4 --- System: LiClOu + C222D in CH CN -45 228 4.39 3.19 i 0.07 7.35 -32 241 4.15 3.26 i 0.10 7.51 -21 252 3.97 3.42 i 0.10 7.87 — 7 266 3.76 3.45 1 0.14 7.95 5 278 3.60 3.50 i 0.16 8.07 28 301 3.32 3.54 i 0.25 8.16 55 328 3-05 --- --- 86 359 2.78 --- --- 76 it '— ox)- ACN NM 6.0} 34: 4}: ‘077 Figure 20. A plot of 1n,K gs l/T for the complexation reactions of Li+ with CZZZD in acetonitrile and nitromethane. 77 ion and is so strongly solvated that considerably more energy must be expended in the desolvation step than for larger ions like Cs+. In this case this desolvation energy is not replaced in the complexation step. Unfortunately, there are not many data available in the literature for lithium- macrocycle complexes to which our data can be compared. Kauffmann gt El- (65) studied Li+-Cle and Li+-0221 complexes in water and report AH = -5.1 kcal/mole and AS = 8 cal/mole-OK for the former and AH = 0.0 kcal/mole and AS = 11.4 cal/mole-OK for the latter. In both cases AS > O and AH becomes more positive for larger ligands. All the Cs-133 chemical shifts measured at various temperatures are listed in Tables 13-15 and shown in Figures 21-23. The formation constants calculated by the KINFIT program are listed in Table 16. All the formation constants listed in Table 16 have been calculated by using the model, éobs = XM5M + XML‘SML + XMA5MA (6) which takes into account ion pair formation as described in Chapter III and Appendix II. In the three systems studied, CsB¢u and CZZZD in pyridine, CSSCN and CZZZD in CHBNO2 and CSSCN and CZZZD in acetonitrile, the curvature in the mole ratio plots is more pronounced as the temperature decreases. This trend implies that the complex formed is stronger at lower temperature. 78 From the calculated formation constants we can see that this is indeed the case. In pyridine the log K = 3.29 i 0.06 at -39°C and it decreases as the temperature increases and at 85°C it has a value of log K = 2.55 i 0.02. In the case of nitromethane the complex formed is weaker and the variation of log K with temperature smaller. At -30°C log K = 1.85 1 0.06 and at 100°C is log K = 1.64 3: 0.06. The Cs+-0222D complex in acetonitrile is also rather weak. At -37°c log K = 2.03 i 0.07 while at 72°C log K = 1.46 5; 0.13. In all three systems studied the complexation reaction is exothermic. Plots of an vs l/T for the above data are shown in Figure 24 and the thermodynamic quantities obtained by linear least squares fitting of these plots are listed in Table 17. In all cases AH < O and AS > O, which indicates that the Cs+-0222D complex is both enthalpy and entrOpy stabilized in all three solvents. Temperature dependence study of the Na+-C222D complex- ation in pyridine and nitromethane was also attempted. But the line widths are extremely broad since sodium has a large quadrupole moment and 0222D provides an unsymmetrical environment around the nucleus. In the case of nitromethane at 5°C the line widths are of the order of 1000 Hz therefore there is a large experimental error involved in the chemical shift measurements and formation constant of the complexes cannot be measured very accurately. 79 00.0 - 00.00- 00.00- 00.00- 00.00- 00.00- 00.00- 00.00- 00.00 00.0 mm.m - 00.00- 00.00- 00.00- mm.00- 00.00- mm.mm- 00.0 00.0 00.0 - 00.0 - 00.00- 00.00- 00.00- 00.00- 00.00- 00.0 00.0 00.0 00.0 - 00.00- 00.00- 00.00- 00.00- 00.0m- 00.0 00.00 00.0 00.m - 00.0 - 00.00- 00.00- 00.00- 00.00- 00.0 00.00 00.0 00.0 - 00.0 - 00.00- 00.00- 00.00- 00.00- 00.0 00.00 00.0 00.0 00.0 - 00.00- 00.00- -- 00.00- 00.0 00.00 00.00 00.0 0m.0 - 00.0 - 00.00- 00.00- 00.00- 00.0 00.00 -- 00.00 00.0 00.0 00.0 - 00.0 - 00.00- 00.0 00.00 00.00 00.00 00.00 00.0 00.0 -- 00.0 - 00.0 00.00 00.00 00.00 00.00 00.00 00.00 -- 00.0 mm.0 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 00.0 000 000 000 000 on 000- 000- 000- 00000 00 .00000000509 0002 .000390000209 050000> pm ms000hzm 00 00000 00 00000000 000 00 00 000.00 000000 00 000000 00000000 000-00 .m0 0000a H-v MU-L O "W" OrN FM ”.0th 0 0-0 - C 0 20.320 I MO CI ! 0. ..ts- fin.» ! I-sFI‘ - ‘If\ 4udv-\ N. u,‘\~ my 0. C V0.00 80 -- 00.00 00.00 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 00.00 00.00 00.00 00.00 00.0 00.00 00.0 0000 000 000- 00- 000 00000 00000 00 .00300009509 0002 00 .00500000509 0002 .00050000Q809 050000> pm 0 0 oz 00 :0 00000 00 00000000 000 :0 As 000.00 20000 00 000000 00000000 000-00 .00 00909 Table 15. C222D in CH CN. 81 Shift of CSSCN (0.01 M) in the Presence of Mole-Ratio-Temperature Data for the Chemical Mole Temperature, 0C Ratio 72° 26° -5° -37° 0.00 -26.88 -3#.64 -39-9l -43.16 0.37 -26.11 -33.71 -38.98 —42.39 0.74 -25.65 -33.24 -38.36 -41.61 1.28 -25.18 -32.31 -37.58 -40.84 1.71 -24.41 —31.54 -36.80 -39-91 2.50 —23.48 -30.92 -36.19 -39.29 2.84 -23.17 -30.61 -35.88 -39.13 3.58 -22.54 -30.30 -35.41 -38.82 4.03 -22.07 -29.98 -35.26 -38.59 5.02 —21.61 -29.52 -34.64 -38.37 82 Table 16. Formation Constants for the Complexation of Cs+ by 0222D in Py, CHBNO2 and CHBCN at Various Temperatures. TOC TOK 103/1 12g_§ 1n K System =CsBPhu + 0222D in Py -39 234 4.27 3.29 3 0.06 7.59 -25 248 4.03 3.14 i 0.03 7.24 -12 261 3.83 3.07 i 0.04 7.07 3 270 3.70 2.96 i 0.02 6.82 25 298 3.35 2.86 i 0.01 6.59 44 317 3.15 2.79 1 0.01 6.41 66 339 2.95 2.64 i 0.01 6.08 85 358 2.79 2.55 1 0.02 5.87 System = CSSCN + 0222D in CHBNO2 -30 243 4.11 1.85 i 0.06 4.25 - 1 272 3.68 1.82 i 0.05 4.20 25 298 3.35 1.79 1 0.03 4.11 77 350 2.86 1.69 i 0.04 3.90 100 373 2.68 1.64 i 0.06 3.77 System = CSSCN + 0222D in CHBCN -37 236 4.24 2.03 i 0.07 4.68 - 5 268 3.73 1.75 i 0.10 4.03 26 299 3.34 1.70 i 0.08 3.92 72 345 2.90 1.46 i 0.13 3.36 83 P Y +40.0 .0.— 7 (PM) +8 5‘0 0.0 .. +66'c . ° - +44 c ' + 25°C 0 +3°c —25°C —39°C —40.o ' ' 2.0 4.0 MozzzlyMc + — ‘ Figure 21. Cesium-133 chemical shifts as a function of 0222D/Cs+ mole ratio in pyridine at various temperatures. 84 N M 60.0 (9 ..., +100°c +77°c 30.0 - +2 5’c ..]°c -30°C 00 1 ' '0.0 M 2.0 4.0 .. c 2220 / M0! Figure 22. Cesium-133 chemical shifts as a function of 0222D/Cs+ mole ratio in nitromethane at various temperatures. 85 AiCtN ("9'“) +72°c -2o.0 - +25%: 9 ‘9 “5°C J —37% -40.0 I J 0.0 2.0 4,0 M0222 VMCs‘ Figure 23. Cesium-133 chemical shifts as a function of C222D/Cs+ mole ratio in acetonitrile at various temperatures. 86 PY 7.0 - 5.0 .- x _c 5.0 _- AACHV NM 4.0 p ' 1 1 3"32.0 3. 4.0 t) 1013 /T Figure 24. A plot of In K.X§ 1/T for the complexation reactions of Cs+ with 0222D in pyridine. acetonitrile and nitromethane. 87 40 04 00000 + 004044 on 0A OA 00.0- 00.0 H 00.4 00.0 H 0.40 20000 04 00000 + 004044 0v 04.0 H 04.0 00.0 H 0.00 N02000 04 00000 + 004044 mm.m- 00.0 H 00.4- 00.0 H 00.4 zomm0 :4 00000 + 20000 00.0- 04.0 H 00.0- 00.0 H 40.0 N020:0 04 00000 + 20000 40.0- 00.0 H 04.0- 00.0 H 00.0 40 04 00000 + 000000 0400\4004 00000 0400\4004 m0 00.04oe\400 00 004040 .mp:0>aom msofipm> 04 00000 40 +00 0:0 +44 40 004000040200 000 004 0040440000 0400040024004 .04 04004 CHAPTER V SPECTROSCOPIC STUDIES OF COMPLEXATION 0F ALKALI METAL IONS, WITH 2,2' BIPYRIDINE IN NONAQUEOUS SOLVENTS INTRODUCTION Studies of the crystal structure of 2,2'-bipyridine have shown that the two pyridine rings are 00p1anar with the N atoms and that the molecule has the transconfiguration (112-114) shown below. Dipole moment measurements of the above compound in solution indicate that the molecule is approximately planar, with the two nitrogen atoms in trans positions about the 2:2' bond (115-117). 2,2'-bipyridine forms strong complexes with many metal ions. The complexes have a metal to ligand stoichio- metry of 1:1. 1:2 or 1:3. In the complexation with a metal or with the hydrogen ions there is chelate ring formation and it is undoubtedly accepted that the ligand has the cis conformation. Most probably the 5-membered chelate ring is 00planar with the rest of the bipyridine molecule. The alkali metals are not expected to form complexes as stable as those of other metals. The predicted trend within the group of the alkali metals is for stability to increase with decreasing atomic number. Quantitative data are not available, but it has been reported that lithium, sodium and potassium complexes of 1:10 phenanthroline, which is quite similar to 2,2'-bipyridine. have been obtained from a methanolic solution but not complexes of rubidium or 88 89 cesium. The complexes Li(Phen) 0104 and Na(Phen)2 0104 are the ones obtained (117). Schilt gt al. (118) tried to prepare the above lithium complex but they obtained a product that appeared to be a mixture of both mono and bis complexes. Grillone gt al. (119) report recently the preparation and isolation of a 1:1 adduct of KBPhu with 2,2'-bipyridine (2.2'BP). The K(2,2' BP)BPhu complex was isolated from acetone solutions which contained 2.2'BP/sa1t ratios of 6:1-7sl. The IR spectrum of the above compound in Nujol was also obtained. Attempts to isolate 2,2'BP adducts of K0104, even by the use of 2.2'BP/salt ratios up to 10:1 were unsuccessful. In general the complexation of 2,2'-bipyridine with alkali metal ions has not been studied in detail. Quite often alkali metal salts are used to achieve high and constant ionic strength in solutions, when complexation of 2,2'-bipyridine with other metal complexes is studied, with the assumption that there is no interaction of the ligand with the alkali metal ions (120). Today we know that alkali metal ions do form stable complexes with many ligands and, therefore, it was of interest to study if there is interaction between 2,2'-bipyridine and alkali metal ions in nonaqueous solvents. For this purpose the reaction of Li+, Na+ and Cs+ ion with 2,2'-bipyridine in different solvents and of Na+ with 2,2'-biquinoline and 4,4‘-dimethyl,2,2'-bipyridine was investigated by using alkali metal NMR and 130 NMR. 90 RESULTS AND DISCUSSION Sodium-23, lithium-7 and cesium-133 chemical shifts were determined as a function of ligand/metal ion mole ratios. In all cases only one p0pu1ation average resonance of the metal ion was observed. It is accepted that the first step in the formation of the metal chelate complex is the formation of the unidentate complex and then it is followed by a rapid chelate ring closure (121). In this reaction the solvent plays an important role. The donicity of the solvent, reflected by its ability to solvate the metal ion, usually expressed by Gutmann's donor number. its dielectric constant and its structure, are quite important factors in determining whether the ligand can successfully compete with the solvent for position in the primary solvation shell. In this study an investigation of the influence of the solvent and of the size of the metal ion as well as the substituent groups on the ligand in the reaction of 2,2'- bipyridine with Li+, Na+ and 08+ was attempted. The reaction of 2,2'-bipyridine with LiBPhu was studied in tetrahydrofuran, methanol, dimethylformamide and propylene carbonate solutions and with LiClOu in nitromethane. All the lithium—7 chemical shifts measured are listed in Table 18. The frequency of the lithium-7 resonance in tetrahydrofuran, methanol and formamide was found to be essentially independent of the ligand/Li+ mole ratio (Figure 25). In these solvents the immediate environment of the lithium ion is not changed upon addition 91 Table 18. Mole Ratio Study of 2,2'-Bipyridine Complexes with Lithium in Various Solvents by Li-7 NMR at 25°C. Solvent Salt [Li+1m_ [2,2'BP]/[Li+] A6(ppm) THF LiBPhu 0.05 0.00 0.93 1.02 1.04 2.00 1.04 3.02 1.04 4.06 0.83 DMF LiBPhu 0.05 0.00 -O.45 0.97 -0.23 2.19 -0.14 2.98 —O.l4 4.00 —0.29 PC LiBPhu 0.05 0.00 0.55 1.05 0.34 2.04 -0.13 3.14 -O.65 ‘ 4.00 -O.97 CHBOH LiBPhu 0.05 0.00 0.43 1.00 0.72 2.04 0.72 3.15 0.67 4.03 0.83 92 Table 18. Continued Solvent Salt ILi+1m_ [2,2'BP]/[Li+] A6(ppm) CHBNOZ LiClOu 0.05 0.00 0.20 0-33 -0-53 0.50 -0.95 0.70 -l.42 1.00 —2.10 1.22 -2.68 1.39 -2.94 1.61 -3.31 1.85 -3.62 2.00 -3.84 2.22 -3-99 2.43 -3-99 2.72 -3-99 3-06 -3-93 3-“4 -3-93 3-81 -3-99 “~58 -3-93 5.02 -3-93 93 ZJDP 8 (P pm)? 1 *L J .THF 5 a g ' MeOH 0.0 l 1 0.0 2.0 4.0 M2 2'BP /ML 3* Figure 25. Lithium-7 chemical shift as a function of 2,2'BP/Li"' mole ratio in THF, MeOH, DMF, and PC solutions. 94 “.0 '— 01431002 0‘) A8 (ppm) -tO 42‘) -4.0 0.0 1.0 20 30 4.0 50 M2 .2' BP/Mu" Figure 26. Lithium-7 chemical shift as a function of 2,2'BP/Li+ mole ratio in nitromethane solutions. 95 of ligand and, therefore, there is no evidence for the formation of a complex. The above solvents have a strong solvating ability (Gutmann donor numbers of 20.0, 25.7 and 26.6 reSpectively) and can compete quite successfully with the ligand for a position in the primary solvation shell. In the case of propylene carbonate (Figure 25), a solvent of intermediate solvating ability (Gutmann donor number 15.1), there is a downfield chemical shift change from the position characteristic of the solvated lithium ion upon addition of the ligand. This shift indicates that there is interaction between Li+ and 2,2'-bipyridine but the complex formed seems rather weak since until a mole ratio of 4.0 there is no indication of approaching the limiting chemical shift. In nitromethane, on the other hand, which is a solvent of weak solvating ability (Gutmann donor number of 2.7) there is a large change of the lithium-7 chemical shift (about 4 ppm) (Figure 26) which indicates a drastic change in the immediate environment of the Li+, from that of the solvated lithium by nitromethane to that of the complexed lithium by 2,2'-bipyridine. From the mole ratio plot of Figure 26 we can conclude that the complex formed is a rather strong one and that it has a stoichiometry of 2:1 since the limiting chemical shift reaches a constant value at a mole ratio of about 2.0. In order to obtain more supporting evidence for the above conclusion, the carbon-l3 nmr spectra of 96 2,2'-bipyridine with LiClOu in the nitromethane at various mole ratios were obtained with the results listed in Table 19 and shown in Figure 27. The Spectrum of 2,2'- bipyridine in nitromethane consists of 4 peaks. The most downfield one was assigned to the 2,2' carbon atoms and the most upfield one to the 5,3,5',3' carbon atoms, with the peaks for the 6,6' and 4,4' carbon atoms in between. Upon addition of LiClOu to the bipyridine solution all the peaks shift downfield and the chemical shift tends to level off after a mole ratio of [Li+]/[2,2'BP] of about 0.5. The magnitude of the chemical shift change is different for the different carbon atoms, the largest one, being for C6,6' and the smallest one for 02,2', while the chemical shift change for the C5,3,5',3' passes through a maximum at a mole ratio of 0.5. The carbon—l3 data strongly support the lithium-7 nmr data that there is indeed a strong Li(2,2'BP)2 complex formed. In order to study further the complexation ability of 2,2'-bipyridine, its reaction with NaBPhu was studied in dimethyl sulfoxide, methanol, dimethylformamide, nitromethane, tetrahydrofuran, pr0py1ene carbonate and acetonitrile. All the sodium-23 chemical shifts and line widths measured in the different solvents are listed in Table 20 and shown in Figures 28-30. In dimethyl sulfoxide, dimethylformamide and methanol, solvents of strong solvating ability, the sodium-23 chemical shift is independent of the [2,2'BP1/[Na+] mole ratio (Figure 28) indicating that at best there is only a very weak interaction between sodium ion and ligand. In 97 Table 19. 13C Chemical Shifts (ppm)a’b for 2,2'-Bipyridine (0.20 M) with LiClOu in Nitromethane. lLi+1/ 502,2' 506,6' 504,4' 505,3,5',3' [2,2'BP] 0.00 150.47 138.21 125.13 121.91 0.30 150.73 139.51 125.88 122.41 (0.26) (1-30) (0-75) (0-50) 0.40 150.83 139.89 126.12 122.55 (0-36) (1.68) (0-99) (0.64) 0.50 150.93 140.28 126.33 122.67 (0.46) (2.07) (1.20) (0.76) 0.61 150.94 140.39 126.39 122.67 (0.47) (2.18) (1.26) (0.76) 0.75 150.91 140.50 126.43 122.63 (0.44) (2-29) (1-30) (0-72) 1.00 150.93 140.46 126.39 122.53 (0.46) (2.25) (1.26) (0.62) 1.20 150.93 140.53 126.42 122.51 (0.46) (2.32) (1.29) (0.60) 2.00 150.86 140.50 126.39 122.44 (0-39) (2-29) (1.26) (0.53) aAll chemical shifts reported are referenced to TMS. _afree bNumbers in parentheses represent the differences5 bipy' 98 01430-02 )- 0 . <5 4.1-3.0 2.0 — A8 ) (ppm) ' . I. 1'? l3 41 4," 4.4' 1.0 — . L I. . . A A “.5353. . I I 12.2. 0.0 l 1 l l l I J 0.0 10 2.0 MLi*/M2.2' BP Figure 27. Carbon-l3 chemical shifts as a function of Li+/2,2'BP mole ratio in nitromethane solutions. 99 Table 20. Mole Ratio Study of 2,2'-Bipyridine Complexes with NaBPhu in Various Solvents by Na-23 NMR at 25°C. Solvent [Na+]M_ [2,2'BP]/[Na+] 0(ppm) Av%(Hz) DMSO 0.075 0.00 1.3 52 0.47 1.0 52 1.00 1.0 55 2.00 1.2 58 3.08 1.0 59 3.95 0.9 64 0H30H 0.075 0.00 4.1 31 0.47 3-9 30 1.00 3.9 30 2.04 3.8 30 2.80 3.6 28 4.00 3.6 30 DMF 0.075 0.00 5.4 37 1.00 5.2 38 2.00 5.2 40 2.90 5.2 40 4.05 5.2 40 CH3N02 0.150 0.00 14.7 12 0.09 13.9 27 0.19 13.1 46 0.27 12.4 61 100 Table 20. Continued Solvent [Na+]M [2,2'BP]/[Na+] 6(ppm) AV%(Hz) 0.36 11.1 83 0.40 10.6 74 0.47 10.1 100 0.63 9.0 124 0.70 8.2 149 0.77 7.8 154 0.83 6.8 181 0.90 6.7 198 1.00 6.8 227 1.13 5.0 1-33 3.6 268 1.44 3.8 325 1.60 1.6 347 2.01 1.1 354 2-30 -0.7 441 2.60 -2.2 432 3-04 -1.8 495 3.30 -3.8 574 3.60 -3.8 605 4.00 —2.8 625 4.70 —4.3 586 5.00 -4.3 600 THF 0.075 0.00 8.1 26 1.01 6.9 56 101 Table 20. Continued Solvent [Na+lm, [2,2'BP]/[Na+] 6(Ppm) Av%(Hz) THF 0.075 2.03 5.0 94 3.03 3.7 104 4.05 3.5 128 PC 0.075 0.00 9.7 64 0.98 8.1 200 2.00 6.4 260 3-04 5-3 352 4.04 4.2 362 CHBCN 0.075 0.00 7.7 11 1.05 5.9 28 2.04 4.8 42 3-03 3-9 52 4.02 3.1 67 102 10.0 1— (P W) ’ o 5.0 5 ' ~ I D M F \\\ ' Econ . e ACZN “—P‘ A fit . D M80 0.0 .4 I 1 0.0 2.0 4.0 M2 2' BF://A4b4°+. Figure 28. Sodium-23 Chemical shift as a function of 2.2'BP/Na+ mole ratio in DMF, CHBOH, THF, PC, CH CN, and DMSO solutions. 3 103 the case of tetrahydrofuran, propylene carbonate and aceto- nitrile, solvents of medium solvating ability (respective Gutmann donor numbers of 20.0, 15.1 and 14.1) there is a downfield chemical shift change and considerable line- broadening of the sodium-23 resonance line upon addition of the ligand. This indicates that there is interaction of the metal ion with the ligand because the immediate environ- ment of the sodium ion Changes and from the symmetric environment of the solvated metal ion it goes to a more unsymmetric one as the increase of the linewidth indicates. The interaction of 2,2'-bipyridine with NaBPhu in the above solvents should be rather weak because until a mole ratio of 4.0 the limiting chemical shift has not been reached. The Change in the linewidths also follows the same trend. In the case of nitromethane a large change in the chemical shift is observed (about 20 ppm), which indicates a drastic change in the environment of Na+, from an oxygen donor like nitromethane, to a nitrogen donor like the 2,2'-bipyridine. Although the chemical shift change is very large, the complex formed should be rather weak because the mole ratio plot starts leveling off at a mole ratio Close to 5.0. The linewidths of the sodium-23 resonance lines follow the same trend (Figure 30). We observe a tremendous change of the linewidth from 12 Hz for the solvated sodium to about 600 Hz for the complexed sodium indicating that 2,2'-bipyridine forms an unsymmetric environment for the sodium ion. From our data we can 104 [SID- 4 CHsNOé HMO - A8 (ppm) SIDE- (MO- -5£)- J I 1 I I 0.0 LO 2.0 p 3.0 . 4.0 5.0 ‘ M22917 Mm“ Figure 29. Sodium-23 Chemical shift as a function of 2,2'BP/Na+ mole ratio in nitromethane solutions. 105 NM .1 000» O O .' .1 AV”: . (H z) 9 o 400» .1 II PC: 1. ll 200» /' T HF ACN 000 . 00 40 20 .&22 BP/MN 0+ Figure 30. Linewidth change of the sodium-23 resonance as a function of mole ratio in.CH3N02, PC, THF and CHBCN solutions. 106 unambiguously conclude that there is a complex formed between Na+ and 2,2'bipyridine since we have such a large change of the sodium—23 chemical shift upon addition of the ligand into the salt solution, and that also the complex formed is a rather weak one. What we cannot unambiguously conclude is the stoichiometry of the complex formed. It is not clear from our data if the complex is 1:1 or a 2:1 as it is the case for the lithium-bipyridine complex. When the above data were analyzed by using the equation for a 1:1 complex, a formation constant of the order of 10 was obtained. Supporting evidence to our assumption that the complex might be a 2:1 is the large Change of the chemical shift which seems to indicate that the Na+ is in a completely different environment than that of the Na+ solvated by nitromethane and that the ligand insulates it from the solvent. In order to see if we can get any clearer picture of the whole situation, the 13C spectra of different mole ratios of [NaBPhu]/[2,2'—bipyridine) in nitromethane were taken with the results listed in Table 21. The chemical shift changes for each of the four carbon signals from the ones of the free ligand as a function of the [Na+]/[2,2'Bipy] mole ratio are shown in Figure 31. The carbon-l3 nmr data agree with the sodium-23 nmr ones. There is considerable change in the chemical shift but not clear-cut evidence about the stoichiometry of the complex. The largest change is for the 6,6' carbons and the smallest one for the 4,4' carbons. UV-visible double 107 Table 21. 130 Chemical Shifts (ppm)a for 2,2'-Bipyridine (0.20 M) with NaBPhu in Nitromethane. ‘Na+]/ 502,2' 506,6' 504,4' 505,3,5',3' [2,2' BP] 0.00 150.47 138.21 125.13 121.91 0.20 150.73 138.68 125.31 122.35 (0.26)b (0.47) (0.18) (0.44) 0.30 150.85 138.85 125.41 122.56 (0.38) (0.64) (0.28) (0.65) 0.50 151.03 139.18 125.56 122.86 (0.56) (0.97) (0.43) (0.95) 0.70 151.14 139.43 125.66 123.03 (0.67) (1.22) (0.53) (1.12) 0.90 151.21 139.53 125.71 123.12 (0.74) (1.32) (0.58) (1.21) 1.00 151.20 139.58 125.71 123.14 (0.73) (1.37) (0.58) (1.23) 1.10 151.23 139.62 125.74 123.13 (0.76) (1.41) (0.61) (1.22) 1.25 151.23 139.66 125.75 123.19 (0.76) (1.45) (0.62) (1.28) 1.70 151.21 139.72 125.77 123.23 (0.74) (1.51) (0.64) (1.32) 108 Table 21. Continued N+ I a 1/ 502,2' 506,6' 504,4' 505,4,5',3' [2,2' BP] 2.00 151.23 139.74 125.79 123.28 (0.76) (1.53) (0.66) (1.37) aAll chemical shifts are referenced to TMS. bNumbers in parentheses represent the differences 6 - 6free bipy 109 z° "‘ 0030102 A8 (ppm) 6.6' 5.3.53 '00 - 2,2' 4.4‘ . / l I 000.0 1.0 . ao MNO‘VMZEBP' Figure 31. Carbon-l3 Chemical shifts as a function of Na+/2,2'BP mole ratio in nitromethane solutions. 110 beam spectrosc0py was also tried but nitromethane strongly absorbs in the whole UV region. The only peak that was obtained was a weak broad one at about 625 nm. This technique does not seem suitable for our study. The complexation of 2,2'-bipyridine with Cs+ was investigated in nitromethane, dimethylformamide, pyridine, propylene carbonate, acetonitrile and methanol by using cesium-133 nmr with all the results listed in Table 22 and shown in Figure 32. There is no evidence of complexation in any of the above solvents. As it is shown in Figure 32, the cesium-133 resonance line is independent of the 2,2'BP / Cs+ mole ratio in all solvents studied. Two substituted bipyridines, the 4,4'-dimethyl,2,2'- bipyridine (4,4' dm, 2,2' BP) and 2,2'-biquinoline (2,2 BQ) shown below 4,4'-dimethyl,2,2'—bipyridine 2,2'-biquinoline were used as ligands to study the effect of the substituent groups on the complexation with Na+. The method used for the study was sodium-23 nmr. The results are listed in Tables 23-24. In the case of the 4,4'-dimethyl,2,2'- bipyridine the only solvent that could be used was 111 Table 22. Mole Ratio Study of 2,2'—Bipyridine Complexes with Cesium in Various Solvents by CS-l33 NMR at 25°C. Solvent Salt [Cs+] [2,2' BP)/[Cs+] Av(ppm) CH3N02 CSSCN 0.05 0.00 61.81 2.06 61.35 2.88 61.19 3.89 60.88 DMF . CSSCN 0.05 0.00 7.51 1.10 7.51 1-97 7-51 2-97 7-51 3-97 7-51 Py CsBPhu 0.015 0.00 49.0 1.73 49.9 3.47 49.9 PC CSSCN 0.05 0.00 41.07 0.92 40.76 2.06 40.45 3.20 40.45 4.16 40.14 CHBCN CSSCN 0.05 0.00 -27.37 0-97 -27-37 2.05 -27.37 3.09 -27.68 4.10 —27.37 112 Table 22. Continued Solvent Salt [Cs+] [2,2' BP]/[Cs+] Ao(ppm) CHBOH CsSCN 0.05 0.00 51.99 1.12 51.68 2.01 51.68 2.96 51-37 3.78 51.06 113 00.0 4 4 :0 N M MeOH :4.:—4 4.;41r 40~ | 0. PV' 400 % 4— 4 a P C 8 (PW) 200 - 0 40 a 4 a DM F 0.0 - 4100 ~ 0.0 2.0 . 4.0 M2 2'8? i/MC 3"” Figure 32. Cesium-133 chemical shift as a function of 2,2' lap/Cs+ mole ratio in CH N0 CH 32’ CHBCN solutions. 30H, Py, PC, DMF and 114 nitromethane. The ligand was not soluble enough in any of the other solvents used in this study. In nitromethane the highest mole ratio we could get was 2.01 for a concentration of NaBPhu of 0.15 M, above that the ligand was insoluble. From Figure 33 we can see that the mole ratio plot up to mole ratio 2.01 that was studied, looks identical to the one obtained with 2,2'-bipyridine and Na+ in nitromethane. These results indicate that the two methyl groups do not change the electronic Cloud around the nitrogen atoms very much. They do provide though a more unsymmetrical environ- ment for the Na+ than the unsubstituted ligand as the line- widths indicate. Biquinoline was only soluble in CH3N02 up to a mole ratio of ligand/salt of 1.34 and in DMSO up to a mole ratio of about 3.0 for a salt concentration of 0.075 M. There is no indication of complexation in DMSO (Figurejfl+) while in nitromethane the solubility limits any conclusions. Chemical shift changes are observed for the solutions studied and all that can be said is that the slope of the mole ratio plot appears to be different than that for the 2,2'-bipyridine itself which seems to indicate that there is some influence of the added benzene rings on the complexing ability of the ligand. 115 Table 23. Mole Ratio Study of 4,4'Dimethyl, 2,2'-Bipyridine Complexes with 0.150 M NaBPhu in CH N02 by 3 Na—23 NMR at 25°C. [4,4' dm,2,2' BPl/ Ad(ppm) AV;(Hz) [Na+) 2 0.00 14.5 13 0.09 13.6 41 0.19 12.5 76 0.26 11.8 83 0-33 11.3 100 0.39 10.7 132 0.43 10.4 130 0.51 10.0 151 0-57 9-5 173 0.63 8.5 198 0.70 8.4 198 0.84 7.0 251 0.94 6.5 298 1.00 6.0 389 1.04 6.2 415 1.14 4.7 466 1.36 3.6 493 1.43 2.4 517 116 Table 23. Continued [40“. dm,2,2' BP]/ A6 0 + 1 real root b2 a3 II T+fi=0+3real I'OO‘tS b2 a3 III .3 + 57 < 0 + 3 real roots Case I, x = A + B Case II and III, use trigonometric form 3 <- g7) Cos¢ = - NIU‘ Cos % I /_ a. 1 o 2 3 Cos (3 + 120 ) 2“. 3 Cos (% + zuo°) Now, solve for [M] in (Al). M II N I balm U Then substitute in following equations, f K CLIMI [ML] = ‘F—_ (L12) K [M] + 1 K. C [M] M KipflVI] + 1 136 6obs : XMGM + xML5ML + xMA5MA (A4) = [M] [ML] [ML] 5obs L 5M + CM 6ML + CM 6MA (45) Use final form of 6 in EQN subroutine. obs Coding symbols in EQN, a = AA p = PP b=BB q=QQ A = AAA r = RR B = BBB ¢ = FE y = R Cos ¢ = CFE CONST(1) = Kip CONST(2) = CM CONST(3) = 51p CONST(4) = 6M u(1) = GML XX(l) = CL u(2) = K: XX(2) = Gobs n glides”? £3: .3... VALQI vav. CON (Q‘MON/ PHNMON/ F hrNS’ OZ HMO—g IIVUVUOA>m20 Nd . 0332—. 0 PI- ONO. -J| UV‘vl -u+442rnflav'uuw4x— n 2222—”. 'UNP- :0 z W --.-M .GWN‘OWMNflO an “’53. unman-uzzannw lflflh¢>uuuu~_ocuxaa a—ornurv- i Z I NOOCom-uo -< D ‘n—nn 2 I N I"? Y‘RHS ST I) ::!)-CONSY(ZD)ICONSYCI’ DRE!!! 2?. men ro 0! this saturlon Is 10 a: uszo rt:.¢- n1). 36cson§§a8§é.;eai:§:I:;.;;i" 'r‘Icr )EOL'O- '3‘9“ "lC0§( 'Q 9"'7.:g§0§7:lfi I-‘I/3o.'.‘COS‘rt/JQOONGLEO'LOO' 9‘” o I ¢"§929f?1 :gfi‘ntvv or 1H: CALCULAYED uttAL cone 0 at (J A” 3 :#11§l¢: cg§¥N£ or re: outsxo: or 0| nun 0|.) < 0 GO' H‘CK ‘ND FIND ANOYHER R007 '1 7“" .6311; 13.no |S untych)A2£.zh pi '?"="' 0 ALL twat! Roots rLuuxcoo) Uns! tut I alt 0 . Ml Igntg . . =3 "2‘"'5'3...f):':?2aés.§:§ 3:!.2Z.§“ 31. ”i “5' '" ”“3"" ' “"3 'rt?vo-m “I'll 2 3 S'JUIY'rCUE uFN . runleuc of {HON f)“ 5 ' ':g¥u.u£.-|l so 70 to an cotrzuuc O h NU '9 23.2.: = |. r “YINUE r N NU i? 98 Y C “VI LIfiIRAlmss 9 MICHIGAN sure umvzasnvl IIHIIHHIHI\llHllNIHIIIWIIHIIIIHIIIIN 3 1 93 03168 8 57