17O nuclear magnetic resonance spectra of steriods

Share Embed


Descrição do Produto

Papers 170 nuclear magnetic resonance spectra of steroids Leland L. Smith, Josef E. Herz, and Edward L. Ezeli Department o f Human Biological Chemistry and Genetics, University o f Texas Medical Branch, Galveston, Texas, USA

The ~70 NMR spectra of cholesterol and31 other steroid alcohols, esters, ketones, and acids enriched by synthesis with 170from H2170have been observed under ordinary operating conditions, and correlations between t70 chemical shift and structure have been adduced. Spectra-structure correlations for these steroids are in conformity with those previously adduced with simpler compounds by others. (Steroids 58:260-267, 1993)

Keywords:cholesterol; sterols; sterol esters; steroid ketones; oxygen-17(170);nuclear magnetic resonance Introduction Interest in applications o f 170 nuclear magnetic resonance (NMR) s p e c t r o s c o p y continues to expand,~-4 but the necessity o f enrichment by synthesis with costly isotope (0.037% natural abundance) and technical aspects of operations (quadrupolar nucleus, I = - 5 / 2 ) limit the study o f important biological compounds. Thus, although applications to amino acids, monosaccharides, and nucleosides have been made, applications to steroids and alkaloids, to drugs and their metabolites, and to oligosaccharides and oligonucleotides inter alia have not been r e p o r t e d : Given the well-known hydrogen bonding and aggregation o f sterols such as cholesterol (cholest-5-en-3flol; 2a) and the recognized utility of 170 spectra for demonstration of hydrogen bonding in other cases, we questioned whether sterol aggregation might be studied via 170 spectroscopy. Spectra of isotopically enriched planar tetracyclic 1,4-chrysenequinone and of semiflexible tetracyclic and pentacyclic l-tetralones at natural abundance have been recorded, 6'7 suggesting that spectra of cholesterol might also be observed. H o w e v e r , we have been singularly unsuccessful in observing the ~70 spectrum of cholesterol at natural abundance, the cholesterol molecule with its tetracyclic ring system and flexible side-chain being considerably larger, therefore with unfavorable molecular motion (correlation time rc), than other tetracyclic c o m p o u n d s for which ~70 studies have been reported. Address reprint requests to Dr. Leland L. Smith, Gail Borden Bldg. 460, University of Texas Medical Branch, Galveston, TX 775550653, USA. Received October 3, 1992; accepted February 8, 1993.

260

Steroids, 1993, vol. 58, June

Nonetheless, despite the limiting aspects o f 170 spectroscopy, we record here that useful ~70 spectra of suitably labeled cholesterol and derivatives may be observed.

Experimental Spectroscopy methods Spectra were recorded on 3.8-35 mM solutions of steroids in CD3CN (Cambridge Isotope Laboratories) at 75 C using a JEOL GX270WB NMR spectrometer operating at 36.6 MHz for 170. An observation frequency range of 40 kHz (850 pm to -250 ppm versus external H~O as reference zero) digitized to 10 Hz resolution (0.1 second acquisition time, 3 mseconds pulse delay, 90° pulse 26/~seconds) allowed accumulation of 5 × 105 acquisitions in a typical overnight (15 hours) period. An extended accumulation was unnecessary in many cases where a high level of J70 labeling was achieved, as few as 105 acquisitions giving good signal-to-noise ratios (S/N) in some instances. Data were processed with an exponential broadening factor of 40 Hz and trapezoidal apodization of 0.1-0.5% as needed to suppress baseline roll from acoustic ringing. Relaxation time T~ (inversion recovery procedure) for [3/3-170]cholesterol was 417/~seconds, that for internal H2170 (at natural abundance) 37.40 mseconds. Spectra were accumulated without ~Hdecoupling and without sample spinning, because better S/N (56:1 versus 52: 1) but no effect of =H-lVOscalar coupling on line shape were observed in coupled spectra. Steroid concentration over a fourfold dilution range had a negligible effect on line narrowing, but the high dilutions caused significant S/N decrease. Chemical shift data are measured versus external H2170 as reference, with an estimated error of -+1 ppm. In some samples, inadvertent inclusion of moisture (from the steroid sample, solvent CD3CN although taken directly from manufacturer's sealed ampule, or moisture imbibition from high-humidity Galveston, TX air) served as an auxiliary internal reference (8o - 3 to -9, Wl:2 120 Hz relative to external H20, depending on amount).

© 1993 Butterworth-Heinemann

170 NMR spectra of steroids: Smith et aL Isotopic syntheses. Syntheses were conducted with [160, 170, tsO]-H20 composed of 20.2% 160, 45.0% 170, and 34.8% tgO from Icon Services, Summit, NJ (costly and no longer readily available) or of 27.1% t60, 21.5% 1'/O,and 51.4% laO from Isotec Inc., Miamisburg, OH (less costly and of greater availability). Labeling of steroids with 170 was accomplished as described hereinafter in detail, the labeled steroids necessarily containing all three oxygen isotopes. However, for simplicity, enriched samples will be identified as [170]-steroids without regard to the presence of the magnetically silent t60 and =sO isotopes. All [170]-steroids were characterized by combinations of melting point (rap) (Kofler), thin-layer chromatography, and IH and 13C NMR spectra in comparison with authentic reference samples. [3~l'/OlCholest-$-en-3~ol (2a). A solution of 200 m g 3tx,5acyclocholestan-6/3-olmethyl ether (lb) in anhydrous 1,2-dimethoxyethane containing 200/zl H2170 and 0.2 ml 6 % trifluoromethanesulfonic acid in 1,2-dimethoxyethane was heated to 50-55 C and stirredfor 3.75 hours in a sealed vialunder N2. Solid N a H C O 3 was added, and the solvent was removed under N2. The residue was dissolved in CH2C12, the solution washed with H20, dried over anhydrous Na2SO4, and evaporated under vacuum. The crude steroid was dissolved in CHCI3, and the CHCI3 solution was applied to two preparative thin-layer chromatoplates. After irrigation with C H C I 3, a total of 127 mg (65.8%) purified [3/31TO]cholesterol (rap 144-146 C), was obtained) Mass spectrometric analysis of the product [3/3-~60,170,IsO]cholesterol showed [M] +, [M + 11+, and [M + 2]+ ions m/z 386, 387, 388, 389, and 390, from whose abundances was calculated the isotope composition 35.9% 160, 38.4% 170, and 25.7% IsO, at variance with the isotope analysis supplied by the commercial supplier of the H2170 (20.2% t60, 45.0% 170, 34.8% 180). W e are unable to account for the discrepancy at this time. Mass discrimination in the ion source, mass analyzer, and ion detector, although possible, does not appear to occur for ions m/z 457-464 and rn/z 367-374 of cholesterol 3/3-trimethylsilylether,9 and we assume mass discrimination for ions m/z 386-390 of cholesterol per se does not account for the matter. The amount of 170 incorporated into a given steroid sample was not further examined, but considerable variation in the amount incorporated occurred (witness the rapidity with which some 170 signals emerged from background relative to others). As few as 105 accumulations were sufficientfor some samples, as many as 1& spectra necessary for others. [3/3-J70]Cholesterol was used to prepare the several sterol esters (2b, 2c, 2d) 5a-cholestan-3/3-ol (3a) and 3a 3/3-acetate, and 5tx-cholestane-3/3,5,6/3-triol(3b) by standard means. 5,6a-Epoxy-Sa-13/]-t'/O]cholestan.3~O-ol (4a). A solution of 35 nag [3/3-1'/O]-2ain 0.5 rnl CH2CI2 was treated with 32 mg mchloroperbenzoic acid in 0.5 ml CH2C12. After 4.25 hours, substrate 2a was consumed, and only product 5a,6a-epoxide (4a) (containing trace of 4b) was detected by thin-layer chromatography. The reaction mixture was neutralized with 10% NaHCO3 solution, and the organic solvent layer was washed with water, dried with anhydrous Na2SO4, and evaporated. The product 5tz,6a-epoxide (4a) was purified by thin-layer chromatography (CHC1Jacetone 9: 1), /trl (CD3CN, 75 C), 0.685 (3H, s, C-18), 0.916 (6H, d, J = 6.59 Hz, C-26/C-27), 0.953 (3H, d, J = 6.23 Hz, C-21), 1.098 (3H, s, C-19), 2.846 (1H, d, J = 4.49 Hz, 6/3H), 3.720 (IH, br m, 3a-H).

5,6~Epoxy-5~[3~t'/Olcholestan-3~ol 3/I-acetate (4c). A mixture of 400 rng KMnO# and 200 rag CuSO4 • 51-120was ground to a fine powder, to which was added 1 m.l CH2CI2, 20 p.l H21"/O, and 74 nag [3/3J'/O]-2c,and then 100 p.1 tert-butanol was added) °

The mixture was stirred for 2 hours at room temperature:in a sealed vial, then diluted with CH2CI2 and filtered through a column of silica gel (for thin-layer chromatography). A fraction containing 20 nag [3/3-t70]-4¢, 8H (CD3CN, 75 C), 0.657 (3H, s, C-18), 0.863 (6H, d, J = 6.22 Hz, C-26/C-27), 0.909 (3H, d, J = 6.23 Hz, C-21), 0.972 (3H, s, C-19), 2.996 (1H, s, 6a-H), 4.725 (IH, br m, 3c~-H), was recovered. Further elution yielded an additional 47 mg [3/3-170]-4¢. Only traces of the isomeric 4a 3jg-acetate were detected by thin-layer chromatography and IH spectra.

5,6~Epoxy-5~-[3~O]cholestan-3/$-ol (4b). A solution of 20 mg [3/3-t70]-4cin 1.2 ml CH3OH containing 0.25 ml H20 and 10 mg LiOH • H20 was kept under N 2 in a sealed vial overnight. After neutralization with two drops of acetic acid, solvents were removed, yielding product [3/3-1~O]-4b.

[tTOl Ketones via ketal hydrolysis ($a, 5b, 5d, 6a, 9, 10a). A solution of 55 mg 17,17-ethylenedioxyandrost-5-en-3fl-ol (prepared from 17-ketone 9b and ethylene glycol in benzene containing toluenesulfonic acid in the usual manner) in 0.7 ml anhydrous dioxane containing 60/zl H2170 was exposed to a trace of dry HC1 gas, It and the mixture stirred at room temperature in a sealed vial overnight. After removal of solvent under N 2, the product was crystallized from anhydrous CH3CN, yielding 40 mg [1', 17-1702]-9b, mp 90 C. By similar means, [170]-ketones 5a, 5b, 5d, 6a, 9a, and 10a were prepared.

[170] Ketones via direct exchange ($a, 6b-6e, 7, 8, 10b, 10c). A solution of 53 nag 5a-3-ketone (5a) in 1 ml anhydrous dioxane containing 50/~1 H2170 was exposed to a trace of dry HCI gas, and the mixture was stirred at room temperature for 4.5 hours in a sealed vial, after which the solvent was removed under N 2. By similar means, [170]-ketones 6b-6e, 7, 8, 10b, and I0c were prepared. For less reactive 6-ketone (7a) and 7-ketone (7b), heating at 70 C overnight was necessary for satisfactory isotope exchange.

[170] Steroid acids by direct exchange (5c, 10d, Ha). A solution of 25 mg 17/3-carboxylicacid (10d) in 1.0 ml anhydrous dioxane containing 50/tl H2170 was exposed to dry HC1 gas and heated at 70 C for 7.5 hours and processed as described. Results After overcoming some unanticipated synthesis problems, it was possible to incorporate 170 from labeled H2170 into steroids and to record their 170 N M R spectra by standard observation means. Spectra o f steroids as small as the C!g-17-ketone 9a and as large as the C27sterol esters 2 b - 2 d were thus readily available for study. In Table 1 are recorded 170 N M R spectra o f 32 steroids obtained under optimum conditions (CD3CN, 75 C, necessarily set by the boiling point 82 C). Useful spectra were also obtained under other conditions, but at the expense o f spectra quality, because 8o, Wv2, and S/N were dependent on solvent and temperature, as expected. For the preferred conditions, Wt/2 ranged from 200 H z for 5d to 937 H z for 7a, thus entirely within a range comparable with that recorded for simpler alcohols such as the isomeric butanols (Wl/2 350-1000 H z ) ) 2 Spectra for the same solution o f [3fl170]cholesterol in CD3CN (106 accumulations) over the

Steroids, 1993, vol. 58, June

261

Papers Table 1 170 NMR data for [170] steroids a Steroid Alcohols [3/~JTO]Cholesterol (2a) 5a-[3/~JTO]Cholesta n-3/3-ol (3a) 5a-[3/3-~70]Cholesta ne-3fl,5,6/3-triol (3b) 5,6a-Epoxy-5c~-[3/3-170]cholestan-3/~-ol (4a) 5,6/3-Epoxy-5/3-[3fl-170]cholesta n-3/~-ol (4b) 5a-[3eJTO]Cholestan-3a-ol (3d) Esters [3fl-170]Cholesterol 3fl-formate (2b) [3flJ70]Cholesterol 3fl-acetate (2¢) [3/3J70]Cholesterol 3fl-propionate (2d) 5a-[3/3-170]Cholestan-3fl-ol 3fl-acetate (3¢) 5,6/3-Epoxy-5/3-[3/3-170]cholestan-3/3-ol 3fl-acetate (4¢) Methyl 3c~-hydroxy-5/3-[3a,24JTO2]cholanate (11bl) Methyl 3e-hydroxy-5/3-[24,24JTO2]cholanate (11b2) Ketones 5a-[3JTO]Cholestan-3-one (5a) 5/3-[3J70]Cholesta n-3-one (Sb) [3-170]Cholest-4-en-3-one (6a) [3JTO]Cholesta-l,4-dien-3-one (6b) Methyl 3-oxo-5/~-[3,24JTO2]cholanoate (Sd) [3,20J702]Preg n-4-ene-3,20-dione (6c) 21-Hydroxy-[3,20J702]pregn-4-ene-3,20-dione (6d) 17a,21-Dihydroxy-[3,20JTO2]pregn-4-ene-3,20-dione (6e) 3fl-Hydroxy-Se-[6JTO]cholesta n-6-one (7a) 3fl-Hyd roxy-5a-[7JTO]cholesta n-7-one (7b) 3/3-Hyd roxy-[7JTO]cholest-5-en-7-one (6a) [7-170]Cholesta-3,5-dien-7-one (Sb) 3/3-Hydroxy-[17J70]androst-5-en-17-one (9a) 3/3-Acetoxy-[1 ',17JTO2]androst-5-en-17-one (9b) 3/3-Hydroxy-[20-170]preg n-5-en-20-one (10a) 3/3,17e-Dihydroxy-[20JTO]pregn-5-en-20-one (10c) 3/3,21-Dihydroxy-[20JTO]pregn-5-en-20-one (1Ob) Acids 3-Oxo-5/3-[24,24J702]cholanic acid (5c) 3/3-Hydroxy-[20,20- 17O2]androst-5-ene-17/3-carboxylic acid (1Od) 3c~-Hydroxy-5/3-[3a,24JTO2]cholanic acid (11al) 3e-Hydroxy-5/3-[24,24J702]cholanic acid (11a2)

$o

Wv2 (Hz)

38.80 38.93 34.13 36.80 36.80 28.50

700 650 527 625 518 350

201.1 198.4 193.1 201.6 193.2 360.0c 38.68 c 358.2 b 141.0 d

525 617 850 484 641 234 300 237 203

555.1 558.3 534.6 504.9 560.8 360.2 b 572.2 534.5 535.6 536.7 558.0 551.6 547.6 547.9 524.1 525.0 363.7 b 569.3 570.9 537.0

475 400 475 488 380 200 342 332 401 696 937 654 765 752 290 390 371 350 342 420

269.9 ~ 270.1 268.2 e 38.68 e 269.0

283 322

342 300 235

a Spectra of samples in CD3CN taken at 75 C. b Ester carbonyl oxygen signal. c 3e-Hydroxyl signal common to [3,24JTO2]-11al and [3,24JTO2]-11b 1 in their admixture. d Methyl (primary alcohol) ester carbinol oxygen. e Assignment of signal to 24-carboxyl group of [24,24-1702]-5c and [3,24-1702]-11al is inferred from apparent partial hydrolysis of the corresponding methyl esters 5d and 11bl during ketal hydrolysis.

range 27-75 C gave the results W~/2and S/N: 75 C, 675 Hz, 52; 73 C, 700 Hz (calculated 706 Hz using the computer calculation program Spectrocalc from Intergalactic Scientific Co.), 33; 64 C, 750 Hz, 35; 54 C, 850 Hz, 29; 45 C, 900 Hz, 31; 36 C, 1,150 Hz, 28; and 27 C, 1,350 Hz, 15. Spectra were also observed using the more viscous CDCI3 as solvent (necessarily at lower temperature 50 C), but with slightly different 8o and somewhat broader signals. For instance, the [3flJ70] cholesterol signal at 8o 39.75 had Wl/2 960 Hz in CDC13 at 50 C; that of its 3/3-acetate (2¢) at 80 194.76 (W1/2 1,140 hz at 50 C, broadened to W1/2 1,620 Hz at 25 C) 262

S t e r o i d s , 1993, v o l . 58, J u n e

and of the 3/3-propionate (2d) at 80 190.38 (Wt/2 1,200 at 50 C, broadened to W~/2 1,400 Hz at 37 C). The ~70 signal of [3/3J70]cholesterol in CC14 at 60 C was 8o 35.7 (Wl/2 900 Hz). Data in Table 1 representing four classes of steroid oxygen functionality (alcohol, ester, ketone, carboxylic acid) conform in detail to related data of others for similar organic compounds of smaller size. The secondary alcohols 2a, 3a, 3b, 3d, 4a, and 4h are characterized by 170 signals in the range 8o 28.5-38.9; esters 2b-2d, 3e, and 4c in the range 80 193.1-201.6; steroid ketones 5-9 and lOa-lOc in the range 80

170 NMR spectra of steroids: Smith et al. Table 2 Ranges of ~O chemical shift data for steroid alcohols, ketones, esters, and carboxylic acids

Function (examples)

Hydroxyl 12a, 3a, 3b, 3(t, 4a, 4b) Ketone (5-9, 10a-10¢) Carboxyl (5¢, 10d, 11a) Ester carbinol (2b, 2e, 2d, 3©, 4¢) Methyl ester carbinol (11b=) Ester carbonyl (St, 9b, 11611

Steroids 8o

Others* Bo

28.5-38.9 504.9-572.2 268.2-270.1 193.1-201.6 141.0 358.2-363.7

15.6-43.8 b 495.3-581.4 240-253c 196-200 d 124-148 346-380

a Data for samples in any nonaqueous solvent are taken from compiled published sourcesfl .=9 Secondary alcohols only, ¢ Simple aliphatic carboxylic acids only; 8o 240-287 for all (aromatic, halogenated, etc.) carboxylic acids. Our data for acetic acid: 8o 271-275 in CD~CN at 75 C (concentration dependent). d Secondary alcohol esters only,

504.9-572.2, and steroid acids 5¢, 10d, and l.la 8o 268.2-270.1, all well within chemical shift and band width limits previously defined (Table 2). Although the steroid ~70 signals are broad, their breadths are nonetheless within the ranges observed for lower molecular weight compounds. The narrowest signals are those of acids 5c, 10d, and l l a (W1¢2235-342 Hz) and of CIg- to C24-esters 5d, 9h, and 11b (Wl/2 200-371 Hz), the narrow signals perhaps reflecting more rapid molecular motion (smaller molecular size, lack of aggregation). Steroid ketones 5, 6a--6d, 9, 10a, and 10b also have narrow signals (Wl/2 290-475 Hz), as does the axial 3a-alcohol (3d) (Wl/z 350 Hz). However, ketones 6e, 7, and 8 display much broader signals (W~/2 654-937 Hz). Whereas the broad signal of 3,20diketone (6e) reflects the merging of individual, unresolved 3- and 20-ketone signals, the broad signals of 6- and 7-ketones 7 and 8 must reflect other matters. Intermolecular hydrogen-bonding possible for the hydroxyketones 7a, 7b, and 8a does not appear to account for signal broadening as the equally broad ketone signal of dienone 8b devoid of hydroxyl group rules out such interpretation. The equatorial alcohols 2a, 3a, 3b, 4a, 4b, 7, and 8a are characterized by broad signals (WI/2 518-937 Hz), suggesting hydrogen bonding. The signals of the steroid esters 2h, 2c, 2d, 3¢, and 4h are also broad (Wl/2484-850 Hz), reflecting an orderly broadening correlated with increased fatty acyl chain length effect on molecular motion. Synthesis of [170]-steroids was achieved by three simple means: (1) by hydrolysis with H2Z70 of 3a,5acyclocholestan-6/3-ol (/-cholesterol) (la) or its 6/3methyl ether lh to yield [3fl-170]cholesterol (2a),8 from which other [3~-)70]-sterols were prepared by standard means; (2) by acid-catalyzed hydrolysis of ketosteroid ethylene ketals to give labeled ketones; H,Is,z4 and (3) by acid-catalyzed exchange of ketone, ester, and carboxylic acid carbonyl oxygen with labeled H2170.15-21 Despite the apparent simplicity of these reactions, there were pitfalls deserving note here, because the

cost of H2170 is high, and availability under present market conditions is uncertain. Initial exploration of the use of microwave radiation (5 minutes, half power level) to promote hydrolysis of cyclosterol lh (or ta) in dioxane to [3/3-170]cholesterol (2a) was not satisfactory, as a prominent by-product formed (probably cholesta-3,5-diene) that necessitated separation from labeled sterol by digitonin precipitation and recovery of [170]-2a from the digitonide. Hydrolysis by heating in the usual fashiong gave a cleaner reaction from which product was isolated by crystallization. Best results were obtained by heating 6fl-methyl ether lb for 5 hours with trifluoromethanesulfonic acid and H2170 in 1,2dimethoxyethane. For labeled steroid ketone synthesis we initially attempted chromic acid oxidation of [3fl-170]-5a-stanol (3a), but the product 5a-3-ketone (Sa) was devoid of detectable isotope. From the acknowledged mechanism of chromic acid oxidation of secondary alcohols via their chromate esters, this loss was not anticipated, 22 but isotope exchange from putative product [3fl-170]-Sa during recovery and purification may have caused isotope loss. In order to conserve costly H2170, an alternative approach was next explored: ketal hydrolysis with H2170 and gaseous HCI in CHaCN solutions according to Turro et al. is,14 involving a mechanism requiring introduction of the oxygen of H20 into product [170]ketones. However, solvent CHsCN was preferentially hydrolyzed, giving rise to spurious 170 signals in spectra of crude steroid ketones in the range 8o 332-338, 309-323, and 271-275 clearly not those of labeled ketones, but here recognized as those of acetamide (CH3CITONHL), acetamide hydrochloride salt or complex (CHaCI'ONH2)x • HCI, and acetic acid (CHsCI702H), respectively. Acid-catalyzed hydrolysis of CH3CN involves initial protonation of nitrile nitrogen (CHsC+~---NH CHsC~N+H), followed by attack of H20 on the protonated species giving an oxygen-protonated species CHsC(O+H2)J~-NH, from which CHsCONH 2 is derived. Other species potentially formed from CHaCN under the action of gaseous HCI include hydrogenbonded CHsCONH 2 dimers, trimers, and oligomers, and imidate hydrochloride CH3C(OC2H4OH)=NH~CIformed from ethylene glycol released during steroid ketal hydrolysis.23-~5 Also, CHaCONH 2 forms salts or complexes with a variety of mineral acids, 24,25 and we recovered such compound (CH3CONH2)x • HCI, nap 122-124 C, 80 318.4, 8c 21.46, 178.27, from CH3CN treated with aqueous HCI (20 mg from 2 ml) and from solutions of CH3CONH 2 so treated. Although stoichiometry was not investigated, composition as an HCI salt or complex was supported by release of CO 2 upon addition to aqueous NaHCO3 solution, by Beilstein test for halogen, by 170 and ~3C spectra different from those of CH3CONH2, and by a mass spectrum (m/z 60, 59 [M] + and 44 [M-CHs] +, 43, 42, and 41 of parent CHsCONH226 but also m/z 62 and 46 not here interpreted). Identification of the spurious 170 signals as those of Steroids, 1993, vol. 58, June

263

Papers CHsCI7ONH2, (CH3CITONH2)x " HC1, and CH3CI702 H was made from 170 spectra of these species observed at natural abundance (CD3CN , 75 C) for the purpose. Concentration-dependent 170 spectra were observed for CH3CONH 2 and CH3CO2H: for CH3CONH 2 saturated solution, 8o 330.5 (W1/2 120 Hz); twofold dilution, 8o 334.6; tenfold dilution, 8o 344.2 (W1/2 120 Hz). For CH3COzH, concentration-dependent signals in the range 8o 267.7-271.0 (Wv2 100 Hz) were observed, with signals from more concentrated solutions at higher field, and with internal moisture observed in the range 8o - 1.6 to -5.3, shifted to lower field with increased concentration, clearly pH-dependent signals. These data indicate that the spurious 170 signals observed at 8o 332-338 are those of odd amounts of contaminant CH3C17ONH2 in the crude steroid ketone products, and that signals at 8o 271-275 are those of CH3C~7OzH also present. By extension, although concentration studies were not conducted for (CH3CONHz)x • HC1, spurious 170 signals in the range 8o 309-323 appear to be those of the HCI salt or complex. Hydrolysis of solvent CH3CN under Turro's conditions 13'14has not previously been recognized. Although further purification of [170]-ketone would remove the spurious [170]-contaminants and the extent of hydrolysis was small, H2170 was consumed in solvent hydrolysis, leaving little to be incorporated more slowly into steroid ketone products. Solvent hydrolysis interference was relieved by using dioxane as solvent, 11under which circumstance 170 was smoothly incorporated into steroid ketones 5-9, 10a, and 10b. Nonetheless, yet another unexpected labeling event occurred under ketal hydrolysis conditions with the keto esters 5d and 9b, the ester carbonyl being labeled by acid-catalyzed exchange at the same time as the steroid ketone carbonyl via ketal hydrolysis. Indeed, ester carbonyl exchange labeling was the predominant reaction with limited H2170; only with increased amounts of H2170 were both functions labeled. Thus, hydrolysis of methyl 3,3-ethylenedioxy5fl-cholan-24-oate in dioxane gave the 5fl-3-ketone 5d with spectrum containing the 3-ketone signal (8o 560.0), but also signals at 6o 360.2 arising from the methyl ester carbonyl group labeled by acid-catalyzed exchange during the reaction and 8o 269.9 from acid 5e formed by ester hydrolysis in the same process. Hydrolysis with lower amounts of H2170 (CH3CN as solvent) gave 3-ketone 5d with spectrum devoid of ketone signal but with signals for the ester carbonyl and carboxylic acid in addition to a signal at 8o 309.5 for the solvent hydrolysis product (CH3C17ONH2)x • HC1. For 3/3-acetate 9b (obtained via ketal hydrolysis in CH3CN), in addition to signals 8o 525.0 of the 17-ketone and 8o 338.1 attributed to CH3CONH2, there was a signal 8o 363.7 also recognized as being within the range of acyl ester carbonyl (8o 346-380z7), reflecting 170 incorporation into the 3/3-acetate carbonyl feature. Thus, both steroid acid methyl ester and steroid alcohol acyl esters are susceptible to acid-catalyzed carbonyl group labeling with H2170. 2£)4 Steroids, 1993, vol. 58, June

To confirm these interpretations as involving JTO exchange phenomena, unlabeled methyl 3a-hydroxy5fl-cholanate (lithocholate) (llb) and 3a-hydroxy-5Bcholan-24-oic (lithocholic) acid (lla) were subjected to ketal hydrolysis conditions with H2170. From keto ester l l b , signals at 6o 360.2 (ester carbonyl of llb) and 8o 269.9 (24-carboxyl of lla) were observed, thus confirming ester carbonyl exchange labeling and ester hydrolysis occurring under these conditions. An alternative origin of the 8o 360.2 signal from [24-170]-ester (llb) reformed from [24-170]-acid (lla) rather than by exchange of ester carbonyl oxygen is most unlikely. From [24,24-1702]-acid (llaz), a single signal at 8o 269.0 was observed, establishing labeling of both carboxyl oxygen atoms, as anticipated. Subsequent methylation of [24,24-1702]-acid (lla2) with diazomethane gave [24,24WO2]-ester (llb2) characterized by two 170 signals 8o 358.2 (ester carbonyl) and 141.0 (methyl ester carbinol), thus establishing the matter definitively. The direct exchange of isotope from H2170 (dioxane, HC1) into ketone, ester, and carboxylic acid carbonyl groups afforded [170]-derivatives suitably labeled, without complications, and is the method of choice for these syntheses. Although acid-catalyzed carbonyl oxygen exchange with H2170 is well recognized for ketones and carboxylic acids and in hydrolyses of acyl esters to acid and alcohol, exchange of acyl ester carbonyl oxygen under like conditions has not been recognized heretofore as a ready means of providing isotope labeling. 14.28 The previously recorded introduction of more than one 180 atom from H2180 into methyl ester [180]-5d in acid-catalyzed hydrolysis of 5d ethylene keta111 may now be regarded as involving lEO incorporation into the 3-ketone and ester carbonyl groups as here established. Because ester carbonyl oxygen exchange appears to be predominant in these reactions of steroid esters, the isotope labeling method may find more general usefulness for labeling ester carbonyl oxygen atoms.

Discussion Data in Table 1 evince that 170 NMR spectra of appropriately labeled steroid alcohols, esters, ketones, and acids may be observed using standard operating procedures. A priori concern for unfavorable molecular motion (correlation time rc) is thus allayed, given adequate levels of 170 isotope incorporation. Moreover, these 170 chemical shift data conform with similar data of simpler organic compounds (Table 2), thus establishing that the tetracyclic nucleus, side chain, and remote functional groups do not alter the general spectral properties of a given labeled functional group. Extension of this approach to other steroids, other classes of natural products, and to drugs and their metabolites now seems worthwhile. The 170 chemical shifts of the 32 steroids of Table 1 show several correlations with structure, some not heretofore recognized in previous studies. The six equatorial secondary alcohols 2a, 3a, 3b, 4a, 4b, and l l a 1 are characterized by chemical shifts in the range

170 NMR spectra of steroids: Smith et al.

60 34.13-38.93, thus entirely comparable with the range 8~ 34.0--39.8 recorded for simpler secondary alcohols.'~ However, hydroxyl conformation appears to influence chemical shift, witness 8o 28.50 for axial 3a-alcohol 3d substantially at higher field (ASo 10.2-10.4 ppm) than equatorial alcohols 2a, 3a, and l l a v This correlation is also in keeping with what few other data there are for rigid epimeric cyclohexyl secondary alcohols. 29'3° Moreover, it is apparent that it is the hydroxyl conformation and not C-5 stereochemistry (or A/B-ring fusion) per se that affects chemical shift, because equatorial 5a-3/3- and 5/3-3a-alcohols 3a and llat have the same chemical shifts, as do the isomeric 5a,6a-epoxyand 5/3,6/3-3/3-alcohols 4a and 4h, whose 3/3-hydroxyl groups are by proton spectra equatorial (axial 3a-hydrogen signal as a characteristic broad multiple0. Nearby substituents may not influence the chemical shifts of 3/3-hydroxysteroids appreciably. There is no effect of the homoallylic AS-doublebond of 2a (80 38.80) in comparison with the 5a-stanol 3a (8o 38.93), but the 3/3-hydroxyl of the 3/3,5a,6/3-triol 3b is weakly shielded (ASo -4.8 ppm) by the 5a- and/or 6/3-hydroxyl, as is the 3/3-hydroxyl of 5,6-epoxides 4a and 4b by the epoxide oxygen (ASo -2.1 ppm) relative to 5a-stanol 3a. However, the comparisons between 5,6-epoxides 4a and 4b and 5a-stanol 3a are not exact ones, because the added A-ring distortion caused by the 5/3,6/3-epoxide feature of 4b obviously must influence matters. Nonetheless, these limited data indicate that structural features (unsaturation, oxygen atom) more than three bonds away from the 3-hydroxysteroid 170 atom do not influence the 170 chemical environment appreciably. Acylation of the 3/3-hydroxyl group of sterols 2a, 3a, and 4h is accompanied by a substantial downfield shift of the carbinol oxygen signal (ASo + 156 to + 164 ppm) as expected, from 8o 36.80-38.93 for 2a, 3a, and 4b to 80 193.1-201.6 for the five esters 2b-2d, 3c, and 4¢. In this case the range of chemical shifts of ester carbinol oxygen is within ranges for secondary alcohol esters of simpler compounds (Table 2). Both ester carbinol and carbonyl oxygens are shielded by increasing carbon substitution in the fatty acyl moiety. Data for 3/3-formate (2b), 3/3-acetate (20, and 3/3-propionate (2d) show a substituent fl-effect of A8o -2.7 ppm and y-effect of Ago -8.0 ppm on the ester carbinol signal. Moreover, ester carbonyl signals of cholanoate esters 5d and l l b are shielded (y-effect Ago -3.5 to -5.5 ppm) in comparison with the carbonyl signal of 3/3-acetate 9h. These/3- and y-effects of carbon substitution correspond closely with those of simple aliphatic esters that are characterized by/3-effect Ago - 3 to - 9 ppm and y-effect Ago - 7 to - 12 ppm (however, in other solvents). 27 The A5-double bonds of 3/3acetates 2¢ and 4¢ exert a minor shielding influence (ASo -3.2 to -8.4 ppm) on the 3/3-carbinol oxygen signals relative to those of the 5a-stanol 3/3-acetate 3e. Our study of steroid ketones involves more diverse structural features; thus, an expanded range of chemical shifts (80 504.9-572.2) was observed. The carbonyl signals of C2r20-ketones 6¢, 10a, and 10b are at lowest

field (80 569.3-572.2); those of six-membered ring ketones 5-8 (excluding dienone 6b) are at mid-range (80 534.5-560.8); and those of the five-membered ring C1917-ketone (9) are at 80 524.1-525.0. The cross-conjugated 1,4-dien-3-one (6b) provides the carbonyl signal at highest field 80 504.9. All these data are well within the range of chemical shifts for cyclic and acyclic ketones of simpler structures. 27 The spectral properties of the steroid ketones are little influenced by the presence of remote functional groups, those of the 17-ketones 9 being uninfluenced by the nature of the remote A-ring 3/3-alcohol or 3/3acetoxyl features, those of the 5/3-3-ketones 5h and Sd little influenced (ASo 2.5 ppm) by the different sidechain features. Moreover, where [170]-ketone and other labeled functional groups are remote from one another (as in ketones Sd, 6e, 9b) both 170 signals occur at frequencies unaffected by one another. Spectra of other steroids doubly labeled in remote sites likewise contain two resolved signals: hydroxyl and ester carbonyl signals for hydroxyester llbl, hydroxyl and carboxyl signals for hydroxyacid llat. However, although the methyl ester l l b z doubly labeled in the geminal ester carbinol and carbonyl oxygens is characterized by 170 signals apparently uninfluenced by one another, other nearby functional groups and carbon substitutions may affect ketone chemical shifts significantly. Comparison of chemical shifts of these steroid ketones with those of simpler aliphatic analogs suggests that the additional carbon substitutions of the steroid ring system shield, deshield, or have insignificant effects. For the latter category, comparisons of chemical shifts of the 3-ketones 5 with those of cyclohexanone (80 558 neat) and with 3- and 4-alkylcyclohexanones, there is little or no significant effect. However, comparison of data for the 20-ketones 6¢ and 10a with the datum for methyl isopropyl ketone (60 557) show deshielding (ASo + 12.3 to + 15.2 ppm) by the additional C- and D-ring carbon (C-12, C-14, C-15, C-18) substituents (8-effects), whereas data for 17-ketones 9 and cyclopentanone (8o 545) evince shielding (ASo -20 ppm) by the same C-12, C-14, C-15, and C-18 carbon atoms now giving rise to y-effects. These limited correlations conform to general conclusions for simpler ketones, where/3- and y-effects of carbon substitutions are shielding, 8-effects deshielding.27 In contrast to insignificant effects of C-5 stereochemistry on chemical shifts of 3-alcohols and their esters, the C-5 stereochemistry may affect 3-ketone chemical shifts, because the 5/3-3-ketones 5h and 5d are deshielded (ASo +3 to +5 ppm) in comparison with 5a3-ketone (Sa). In simple alkyl-substituted cyclohexanones, equatorial 3-methyl groups exert insignificant effects, and axial 3-methyl groups deshield (ASo +7 to + 10 ppm) the carbonyl. 31 Other nearby structural features of steroid ketones also exert significant effects on the carbonyl group. The carbonyl of A4-3-ketone (6a) is substantially shielded (ASo -20.5 to -23.7 ppm) in comparison with that of saturated 3-ketones $s and Sb, but the carbonyl of AS-7-ketones 8 is shielded much less (ASo -3.7 to

Steroids, 1993, vol. 58, June

265

Papers - 4 . 0 ppm) r e l a t i v e to that o f 5o~-7-ketone (7a). S i m i l a r

shielding effects (A~o - 1 0 to - 3 0 ppm) of a,fl-unsaturation are observed for simple cyclohex-2-enones. 27 Whereas extension of conjugation from a,fl-enone to a,fl;y,8-dienone does not affect the ketone chemical shift (compare 8a and 8b), the cross-conjugation of the a,fl;a',fl'-dienone (6b) shields the carbonyl very substantially (A~5o -29.7 ppm), a matter not heretofore recognized. This distinction offers yet another ready means of differentiation between A4- and A 1,4-3-ketosteroids. Other effects of structure on steroid ketone spectra are suggested by comparisons of spectra of the three 5c~-ketones 5a and 7. The carbonyl of 5a-7-ketone (7b) is shielded relative to those of the analogous 5a-3-ketone (5a) (A~5o - 3 . 5 ppm) and isomeric 5a-6-ketone 7a (ASo - 6 . 4 ppm). However, an opposite relationship was observed in spectra of the analogous a,fl-unsaturated A4-3-ketone (6a) and AS-7-ketone (8), with the AS-7-ketone carbonyl significantly deshielded (At5o +13.0 to +13.3 ppm) relative to that of M-3-ketone (6a). These effects appear to reflect the sum of effects of vicinal carbocyclic substitution, ring fusion stereochemistry, and ring conformation. Too few examples are available to test these matters further. The 20-ketone carbonyls of a-ketols 6d, 6e, and 10b are substantially shielded (A~o -34.5 ppm, average) by the 2 l-hydroxyl in comparison with those of parent 20-ketones 6c and 10a and of the isomeric a-ketol 17ahydroxy-20-ketone (10c). The shielding is attributed to intramolecular hydrogen bonding well known for 21hydroxy-20-ketones. Moreover, because the 170 signals of 21-hydroxy-20-ketone (6d) and 17a,21-dihydroxy-20-ketone (6e) are not significantly different nor are those of 20-ketone (10a) and 17a-hydroxy-20-ketone (10c), it is concluded that there is no hydrogen bonding between 17a-hydroxyl and 20-ketone groups. Although the 3- and 20-ketone signals of A4-3,20diketone (6e) are resolved (Figure l), those of the 21hydroxy- and 17o~,21-dihydroxy-A4-3,20-diketones 6d

I 700

'

'

'

'

I 600

'

'

'

'

I 500

' ' ' '

I 400

'

'

'

'

I 300

'

'

,

'

I 200

'

'

''

I

'

100

'

'

'

|

'

'

'

'

Table 3 R6sum~of effects of steroid structure on 170 chemical shifts Oxygen function

Structural feature

Effect (Ago, ppm) a

3-Hydroxyl (equatorial) 3fl-Hydroxyl

Inversion to axial A5-double bond 5e-hydroxyl 5,6-epoxide Acylation AS-double bond 5fl,6/3-epoxide C-substitution

-10.1 to -10.4 -0.13 -4.8 -2.1 + 156.4 to + 164.3 -3.2 -8.4 fl-effect -2.7 1,-effect -8.0 y-effect -3.5 to -5.5 +3 to +5 -20.5 to -26.2 -29.7 y-effect -20 g-effect +12.3 to +15.2 -32.3 to -36.6

Ester carbinol

Ester carbonyl

C-substitution

5cx-3-Ketone A4-3-Ketone Ketone

C-5 inversion a,/3-unsaturation Cross conjugation C-substitution

20-Ketone

21-hydroxylation

a Positive ASo values are deshielding; negative Ago values are shielding.

and 6e are not. This degeneracy arises fortuitously from substituents vicinal to each carbonyl, a,fl-unsaturation deshielding the 3-ketone, the 21-hydroxyl shielding the 20-ketone, to yield a single signal unresolved at the 6.34 T field of our spectrometer. The assorted effects of structure on chemical shift for the steroids here examined are summarized in Table 3. These observed effects are those of carbon and oxygen substitutions at three or more bonds from the '70 substituent, but each effect corresponds to related effects previously observed by others with much simpler compounds. It remains to be determined what effects other less remote carbon and heteroatom substituents may have on spectra of 170-steroids. No effect of oxygen isotope on 13C signal of the carbinol carbon of cholesterol was detected, neither by chemical shift nor by line width suggesting isotope effect. However, the field of our spectrometer and the limited extent of isotope enrichment of the [170]-sterol preparation appear to be inadequate for detection of small isotope effects in 13C spectra. Moreover, any observed isotope effect in the 13C spectrum of the [170]sterol may be the sum of 1A13C(170) and IAI3C(nSO) effects that may possibly be of opposite signs, thereby canceling one another. One-bond isotope effects of 180 on 13C (1A13C(180)) generally involve upfield (shielding) shifts. 32 Analogous 1An3C(170) effects in 13C spectra of alcohols have not been studied, but a downfield (deshielding) 1A13C(170) effect (-0.20 ppm) of 170 in the 13C spectrum of acetone appears to be the case. 33

0

Acknowledgments Figure I 170 NMR spectrum of pregn-4-ene-3,20-dione (progesterone) (6c} showing resolution of the 20-ketone signal (at low field) and A4-3-ketone signal. The minor signal at high field is that of moisture inadvertently present in the sample.

266

S t e r o i d s , 1993, vol. 58, J u n e

The assistance of Dr. Khingkan Lertratanangkoon, Department of Pharmacology, University of Texas Medical Branch, in recording mass spectra is gratefully ac-

~70 NMR spectra of steroids: Smith et al. knowledged, as is the financial assistance of the John Sealy Memorial Endowment Fund, University of Texas Medical Branch.

18.

References 1.

2.

3. 4. 5.

6. 7.

8.

9. 10. 11. 12. 13.

14. 15.

16. 17.

Kintzinger J-P (1981). Oxygen NMR. Characteristic parameters and applications. In: Diem P, Fluck E, Kosfield R (eds), NMR Basic Principles and Progress. Springer-Verlag, Berlin, Vol. 17, pp. 1-64. Kintzinger J-P (1983). Oxygen-17 NMR. In: Laszlo P (ed), NMR of Newly Accessible Nuclei. Volume 2. Chemically and Biochemically Important Elements. Academic, New York, pp. 79-104. Boykin DW, Baumstark AL (1989). 170 NMR spectroscopy: assessment of steric perturbation of structure in organic compounds. Tetrahedron 45:3613-3651. Boyken DW (ed) (1991). 170 NMR Spectroscopy in Organic Chemistry. CRC Press, Boca Raton, FL. Woodward RW (1991). tTO-NMR as a mechanistic probe to investigate chemical and bioorganic problems. In: Boykin DW (ed), 170 NMR Spectroscopy in Organic Chemistry. CRC Press, Boca Raton, FL, pp. 115-139. Baumstark AL, Dotrong M, Stark RR, Boykin DW (1988). 170 NMR spectroscopy: origin of deshieiding effects in rigid, planar molecules. Tetrahedron Lett 29:2143-2146. Boykin DW, Dewprashad B, Eisenbraun EJ (1990). t70 NMR studies on alkyl-substituted 1-tetralones: effect of torsion angle change and repulsive van der Waals interactions. J Org Chem 55:425-429. Hudgins LC, Parker TS, McNamara DJ, Bencsath FA, Field FH (1988). Limits for the use of (180)cholesterol and (lSO)sitosterol in studies of cholesterol metabolism in humans. Biomed Environ Mass Spectrom 17:463-470. Lee WNP, Byerley LO, Bergner EA, Edmond J (1991). Mass isotopomer analysis: theoretical and practical considerations. Biol Mass Spectrom 20:451-458. Syamala MS, Das J, Baskaran S, Chandrasekaran S (1992). A novel and highly B-selective epoxidation of AS-unsaturated steroids with permanganate ion. J Org Chem $7:1928-1930. Anderson WG, Byon CY, Eck CR, Gut M (1984). Labeled bile acids. I. Synthesis of 180-methyl hydroxy-5B-cholan-24oates. J Label Compounds Radiopharm 21:59-64. Christ HA, Diehl P, Schneider HR, Dam H (1961). Chemische Verschiebungen in der kernmagnetischen Resonanz von 170 in organischen Verbindungen. Heir Chim Acta 44:865-880. Turro NJ, Paczkowski MA, Wan P (1985). Magnetic isotope effects in photochemical reactions. Observation of carbonyl 170 hyperline coupling to phenylacetyl and benzoyl radicals. 170-enrichment studies. J Org Chem 50:1399-1404. Kabalka GW, Goudgaon NM (1990). 170-Enrichment methods. In: Boykin DW (ed), 170 NMR Spectroscopy in Organic Chemistry. CRC Press, Boca Raton, FL, pp. 21-37. Lawson AM, Leemans FAJM, McCloskey JA (1969). Oxygen18 exchange reactions in steroidal ketones. Determination of relative rates of incorporation by gas chromatography-mass spectrometry. Steroids 14:603-615. Wilton DC (1976). Is a Schiffbase involved in the mechanism of the A4-3-oxosteroid 5a- or 5B-reductases from mammalian liver? Biochem J 155:487-491. Vouros P, Harvey DJ (1980). The electron impact induced

19. 20.

21.

22. 23. 24. 25.

26. 27.

28. 29.

30.

31. 32.

33.

cleavage of the C-17-C-20 bond and D-ring in trimethylsilyl derivatives of C21 steroids. Reciprocal exchange of trimethylsilyl groups. Biomed Mass Spectrom 7:217-225. Bevins CL, Bantia S, Pollack RM, Bounds PL, Kayser RH (1984). Modification of an enzyme carboxylate residue in the inhibition of 3-oxo-AS-steroid isomerase by (3S)-spiro[5a-androstane-3,2'-oxirane]- 17fl-ol. Implicationsfor the mechanism of action. J Am Chem Soc 106:4957-4962. Eriksson CG, Eneroth P (1987). Studies on rat liver microsomal steroid metabolism using 1SO-labelledtestosterone and progesterone. J Steroid Biochem 28:549-557. Gotto J, Miura H, Inada M, Nambara T, Nagakura T, Suzuki H (1988). Studies on steroids. CCXXXVIII. Determination of bile acids in liver tissue by gas chromatography-mass spectrometry with negative ion chemical ionization detection. J Chromatogr 452:119-129. Murphy RC, Clay KL (1990). Preparation of labeled molecules by exchange with oxygen-18 water. In: McCloskey JA (ed), Methods in Enzymology. Vol. 193. Mass Spectrometry. Academic, San Diego, pp. 338-348. Kirk DN, Hartshorn MP (1968). Steroid Reaction Mechanisms. Elsevier, Amsterdam, pp. 30-33. Beckwith ALJ (1970). Synthesis of amides. In: Zabicky J (ed), The Chemistry of Amides. Interscience, London, pp. 73-185. Homer RB, Johnson CD (1970). Acid-base and complexing properties of amides. In: Zabicky J (ed), The Chemistry of Amides. Interscience, London, pp. 187-243. Challis BC, Challis JA (1979). Amides and related compounds. In: Barton D, Ollis WD (eds), Comprehensive Organic Chemistry. The Synthesis and Reactions of Organic Compounds. Vol. 2. Nitrogen Compounds, Carboxylic Acids, Phosphorus Compounds. Pergamon, Oxford, England, pp. 957-1065. Stenhagen E, Abrahamsson S, McLafferty FW (1974). Registry of Mass Spectral Data. Wiley, New York, Vol. 1, p. 12. Boykin DW, Banmstark AL (1991). t70 NMR spectroscopic data for carbonyl compounds: I. Aldehydes and ketones. II. Carboxylic acids and derivatives. In: Boykin DW (ed), IZO NMR Spectroscopy in Organic Chemistry. CRC Press, Boca Raton, FL, pp. 205-231. Ott DG (1981). Syntheses with Stable Isotopes of Carbon, Nitrogen, and Oxygen. Wiley-Interscienc¢, New York. Chandrasekaran S (1991). tTO NMR spectroscopy of single bonded oxgyen: alcohols, ethers, and their derivatives. In: Boykin DW (ed). lzO NMR Spectroscopy in Organic Chemistry. CRC Press, Boca Raton, FL, pp. 141-203. Eliel EL, Liu K-T, Chandrasekaran S (1983). t70 NMR spectra of equatorial and axial hydroxycyclohexanes and 5-hydroxy1,3-dioxanes and their methyl ethers. Org Magn Reson 21:179-182. Crandall JK, Centeno MA, BCrresen S (1979). Oxygen-17 nuclear magnetic resonance. 2. Cyclohexanones. J Org Chem 44:1184-1186. Risley JM, Van Etten RL (1990). Properties and chemical applications of 1sO isotope shifts in 13C and 15N nuclear magnetic resonance spectroscopy. In: Diehl P, Fluck E, Gfinther H, Kosfeld R, Seelig J (eds), NMR Basic Principles andProgress. 22. Isotope Effects M NMR Spectroscopy. SpringerVerlag, Berlin, pp. 81-168. Eichoff U, Hull EW, Scherbakov VA, Khlopin VG (1979). Isotopic effects in the 1H-, 13C- and 170-magnetic resonance spectra of water molecules in acetone. In: Kundha E, Lippmaa E, Saluvere T (eds), Magnetic Resonance and Related Phenomena. Proceedings of the 20th Congress Ampere. SpringerVerlag, Berlin, p. 460.

Steroids, 1993, vol. 58, June

267

Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.