Edward J. Drexler and Kurt W. Field

I An NMR Study of Keto-Enol University of Wisconsin-Whitewater

Whitewater. 53190 Tautomerism in ,8-Dicarbonyl Compounds

Ketones readily undergo acid or base catalyzed a-hydro- gen substitution reactions in which the rate determining step is the formation of the enol or enolate anion

H n

Compared to the acid catalyzed process, the self enolization (tautomerization) of most ketones is negligible;' the keto form is favored almost exclusively (I ). &Dicarbonyl com- pounds also undergo tautomerization; however, the en01 tautomer may predominate a t equilibrium. The enol iso- mer is stablized relative to the keto form via resonance through the conjugated *-system and by intramolecular hydrogen bonding.

0 0 I IH I I =

R',C\C/C\R,,

\R

Ke to

(1) 1n1

En01

An investigation of the keto-enol tautomerization of 0- dicarbonvl compounds provides an excellent underprad- uate physwd organic o r instrumental analvsis experiment. The eiiwts of ;<divnt, omrentration (1, :j,, and tempera- ture (4) on the equilihrium constant can be shown and the thermodynamic functions for enolization (4) obtained.

A procedure for the quantitative determination of the equilihrium constant (Keq) for the tautomerization of 0- dicarbonyl compounds was developed by Kurt Meyer (5); the method involves the direct titration of the enol with bromine. The equilibrium constants have also been ascer- tained by infrared (4, 6 ) , ultraviolet (71, and nuclear mag- netic resonance techniques (8-11 ), and by refractive index measurements (4, 12). The volumetric method may, in

' Acetone exists as 0.00025% end; see Ref. (I). 2 With 8-dicarhanvl eom~ounds in which R' and R" (eqn. (2))

are not idkntical, two different enols may be possible, (I) and (11). Due to differences in Rand R" (st+ and electronic) the two ends should have different equilibrium constants. The determination of these rate constants is usually not possible by proton nmr tech- niques because the proton and electronic shifts are faster than the proton nmr time scale. The Koqk can he obtained for these suh- strates via orveen-17 nmr (14). infrared (15). and ultraviolet (16) . - spectroscopy.

392 / Journal of Chemical Education

Percent Enol Tautomerr and Equilibrium Connants far O-Dicarbonvl Com~ounds

Acety lacetone Ethyl Aceto-

acetate Ethyl Ben-

roylacetate l-Benroylace-

tone Diethyl 1.3-

Acetone-

ate

Neat

Enol. % Keq"

74.7 2.95

20% in CCI,

Enol. % K e q a

90.3 9.33

OKeq = lenol)f(ketal:rPectra recorded at 37 t 2"c be = end; K = keto; t =total.

.e 'b l,., .a 7s " ,. * - m ACETYLACETONE ... , .

-.~ -

OFFSET h, C ' C ~ 660 H I I

I CH, I

4- 30 I,, .o so .a * - ETHYL ACETOACETATE . '

k TMS CH3

e~~ 1 .a -L

OFFSET 437 H Z

I I I 1 .a I. I,', .. %o %L

Figure 1. Proton nmr spectra of acetyiacetone and emyl acetoacetate as neat lbquids at 37 + 2'C

some instances, give erroneous results due to the chemical disturbance of the equilihrium. A quantitative determina- tion of K., by absorption methods may be complicated by deviations from Beer's Law and possible differences in the absorptivities of the keto and en01 carhonyl hands. Nmr appears to be the method of choice because under normal

operating conditions, the conversion rate between the two tautomers is sufficiently slow so that a combined spectrum of each tautomer is ~ h t a i n e d . ~ In addition to providing in- formation ahout K,, the nmr method can he used to deter-

->

mine the thermodynamic parameters for tautomerization bv measurine the temperature dependence of K,.. --

We offer herein experiments designed to use the nmr ~Dectrometer as an analvtical instrument from which an understanding of a number of the factors involved in the keto-enol tautomerism is obtained.

Determination of Enol Content

In the initial portion of the experiment, the enol content of several 0-dicarbonyl compounds and their equilibrium constants are determined (see the table). The spectra of acetylacetone and ethyl acetoacetate and their interpreta- tions are reproduced in Figure 1.

Solvent Effects The second portion of the experiment involves the deter-

mination of the equilibrium constants in several solvents a t various concentrations (9, 10) and constant temperature. Figure 2 reveals that both solvent and concentration have a pronounced effect on K.,. The en01 form of a P-diketone or 6-keto ester is the least polar of the two tautomers because the intramolecular hydrogen bonding in the enol helps re- duce the dipole-dipole repulsion of the carhonyl groups which occurs in the keto form (I). Consequently, as the polar carbonyl compounds are progressively diluted with non-polar solvents the enol content of the system increases. Likewise, a t constant concentrations, the differences in po- larity between these "inert" solvents also produces predict- able shifts in the equilibrium. Conversely, progressive dilu- tion of the 0-dicarbonyl with a more polar solvent than the solute increases the stability of the keto form relative to the enol.

Thermodynamic Functions

The final portion of the experiment involves the deter- mination of the thermodynamic parameters for enolization by measuring Keq as a function of temperature (4.10). Fig- ure 3 represents a plot of log K , versus 1/T for the tau- tomerization of acetylacetone. The thermodvnamic data for acetylacetone determined hy this method-is: AH(-2.7 kcallmole), AS(W.3 callmole deg), and AGd-776 cal).

Experimental Section

The nmr spectra were recorded on a Varian A-60 instru- ment eau i~oed with a variable temperature probe. The probe tekperatures were determined b y using ethylene glycol reference. The spectra in Figure 2 were ohtained using tetramethylsilane (sealed capillary) as an internal reference. The solvents (s~ectro grade) and B-dicarhonvl compounds were used as received kxcept for acetylacetone and ethyl acetoacetate which were distilled prior to use. For the solvent studies, weight percent solutions of the so- lutes (2,5, 10,20,40,60, and 80%) were prepared.

The equilibrium constantsVthe table), solvent data (Fig. 2), and thermodynamic parameters (Fig. 3) were obtained by integrating (10 determinations) the appropriate en01 and keto signals. The standard deviation in the determina- tion of percent en01 was less than f 4%.3 The slope of the line from a plot of log K., versus 1/T afforded -AHIR. The free energy changes at the various temperatures were cal- culated using AC = -2.303 RT log K.,.

Proton Assignments

The proton assignments for the resonances of the com- pounds listed in the table are based on relative signal in- tensities, chemical shifts, and peak multiplicities. In many

20 40 60 80 100 WEffiHT PERCENT

Figure 2. Effect of solvent and concentration on the ketwno l equilibrium of acetylacetane at 37 + P C . (0) CCI,; (A) CHCI,: (0) CH2CI2: (A) HCON(CH,),.

Figure 3. The temperature dependence of K., for the tautomerization of acetylacetone.

cases the assignments may be confirmed by diluting the 0- dicarhonyl with a non-polar solvent which increases the enol content, or by measuring Keg a t two temperatures, the higher temperature favoring the keto form.

Additional Experiments

The literature ahounds with other keto-enol systems amenable to nmr study. The method has been used to ex- amine a homologous series of 0-diketones in which the structures of R' and R" (eqn. (2)) are varied. Such studies revealed that the en01 content increases as the steric re- quirements of the suhstituents increase (13). In addition to variations in Keg, solvent effects are also responsible for predictable chemical shifts of the -OH proton in the 0- dicarbonyl relative to the neat material (9).

Literature Cited

11) Could, E. S.. "Mechanism and Structure in Organic Chemistry." Holf, Rimhart. and Winrtnn,NewYork, 1 9 5 9 , ~ . 376ff.

(2) Ward, C. H., J. CHBM. RDlIC..39.9S (1982). (3) Lcxkwod. K.L.. J. CHEM.EDUC..42.481 11965). (4) Dawber, J. C.,and Crane. M. M...J.CHEM. BDUC.44. IS0 11967). (5) M ~ ~ ~ ~ , K . H . , A ~ ~ . . 3~n.21211911) . 16) Porlin.,J..and Rernsiein. H. ,J Amer Chem. Sor.. 73.4353 (1951). 17) Rrie~leh.G..andStrohmeier. W . . Z Noturlnrsch.. 6b.6 119511. 18) Jarnett, H. S . , S a d l ~ i , M. %and Shnnlery. .J. N.. J. Chem Phys.. 21,2W2(1953). 19) RII@E, M.T..and Burdeit. J. L.. Con. J Chem.. 43.1516 119651.

(10) Alien. G.,snd Dwek, R.A., J. Chcm Soc, 8. 161 11966). 111) Sehweitzer.G. K..and Benson. E. W.. J Chem Ew Doto. 13.452 (1SfiRi. (12) Meyor. K. H., and Wilisn. F. G.. Them R r r . 47,837 11914). 119) Kmhimura. H..Saito. J..and Okubo.T.HuJI. Chem Snc. J o g . 46,692 119731. (14) Tnm,deL1ky,M..Luz,Z..and Marur.Y.. J. Amer Chrm Soc. R9.1183 (1967). (16) Fomon, S., Morenyi,F..and Niisson, M.,Arlo Chem Srand., 18,1208, 11964). (16) Rarnum, D. W.. J~lnorp. Nucl Chem.. 21. 221 (1961).

Computer programs are available from the authors to calculate the K,,, % end and standard deviations, and to determine the thermodynamic quantities. They are written in PL1 language.

Voluk53, Number 6, June 1976 / 393