copper analysis in type metal by atomic absorption gave very good agreement with results obtained by the neo-cuproine extraction method (6) as seen in Table VI. It is recommended that, when analyzing for impurities, a 2.5-gram sample be dissolved and diluted t o 250 ml t o make a more concentrated sample, and then use conditions listed in Table 11. Table VI1 shows the effects of tin, antimony, and lead on copper in fluoboric acid. I n summary, analysis of the major constituents and trace impurities of type metal can be done by atomic absorDtion with no special preparation or treatment except for preparation of standards when samples are dissolved in the fluoboric(6) A. R. Gahler, ANAL. CHEM., 26, 577 (1954).
nitric acid mixture. The atomic absorption results compare very well with standard wet chemistry results and in most cases the procedure is quicker and simpler. ACKNOWLEDGMENT
This paper has been approved for publication by A. N. Spence, Public Printer of the United States. The work was done in the Tests and Technical Control Service under the direction of George G. Groome, Technical Director. RECEIVED for review September 20, 1971. Accepted December 13, 1971. Mention of commercial products does not imply endorsement by the U.S. Government Printing Ofice.
Phototautomerism in the Lowest Excited Singlet State of 4-Methylumbellife~rone G . J. Yakatan,' R. J. Juneau, and S. G . Schulman College of Pharmacy, UniGersity of Florida, Gainesville, Flu. 32601 COUMARIN IS WIDELY distributed in nature and some of its derivatives are of great importance in chemistry and medicine. Many coumarins are naturally fluorescent. Goodwin and Kavanagh (1, 2 ) measured the fluorescence of several coumarin derivatives as a function of pH. Fluorescent indicators have been used for acid-base titrations. For example, Chen (3) indicated that a coumarin derivative, 4-methylumbelliferone, was a useful indicator to follow the p H change induced in the carbonic anhydrase catalyzed hydration of Cor. Several workers have utilized fluorogenic coumarin substrates in enzyme determinations. Robinson ( 4 ) measured p-glucosidase activity by the fluorescence of 4-methylumbelliferone released by action of the enzyme on 7-(/3-~-glucopyranoxyloxy)-4-methylcoumarin. Mead et al. ( 5 ) measured the activity of the same enzyme o n the galactoside substrate. Similarly, sulfatase has been measured by its action on the sulfate ester of 4-methylumbelliferone (6, 7) and alkaline phosphatase by its activity on the phosphate ester (8). The numerous applications of the fluorescence properties of the 7-hydroxycoumarins (umbelliferones) has led to much recent interest in the description of the fluorescence of these molecules in solution (3, 9-11). Creaven and coworkers (9) reported that 7-hydroxycoumarin showed excited state ioniza-
tion from pH 1 to 9. Below p H 9, the excitation shift to shorter wavelength was attributed to the change in the ground state, from the anion to the non-ionized molecule, but the fluorescence emission remained that of the ionized species. These same authors found it difficult to explain the fluorescence shift to longer wavelength at p H 1 and postulated that the fluorescing species at p H 0-1 could be a dimer hydrogenbonded through the free 7-hydroxyl group (9). Fink and Koehler (10) studied the fluorescence of 7hydroxycoumarin in the p H range 1.2 to 11.2. They found a n intense blue fluorescence band which was p H independent for all pH's greater than 2.2. At p H values between 1.2 and 2 . 2 , the latter emission red shifted and decreased in intensity. These authors assigned this phenomenon to an excited-state acid-base equilibrium since the ground state pK, of 7-hydroxycoumarin was approximately 8 (10). Here, the postulate that the lower energy emission was due to the unionized molecule while the higher energy blue fluorescence was due to the 7-hydroxycoumarin anion, indicated that the nature of the fluorescence of coumarin and its derivatives was still in question. This, coupled with our interest in excited state equilibria led us to reinvestigate the nature of the fluorescence of COUmarin and some of its derivatives.
Present address, College of Pharmacy, University of Texas, Austin, Texas 78712.
EXPERIMENTAL
( I ) R. H. Goodwin and F. Kavanagh, Arch. Biochern. Biopliys., 27, 152 (1950). (2) Zbid.,36, 451 (1952). (3) R . F. Chen, A m / . Lett., 1 , 4 2 3 (1968). (4) D. Robinson, Biocliem. J . , 63, 39 (1956). (5) J. A. R . Mead, J. N. Smith, and R. T. Williams, ibki., p 39. (6) W. R . Sherman and E. F. Stanfield, ibid.. 102, 905 (1967). (7) G. G . Guilbault and J. Hieserman, ANAL. CHEM.,41, 2006 (1969). (8) G. G. Guilbault et a/., A t i d . Lett., 1, 333 (1968). (9) P. J. Creaven, D. V. Parke, and R. T. Williams, Blochem. J., 96, 390 (1965). (10) D. W. Fink and W. R. Koehler, ANAL.CHEM., 42, 990 (1970). (11) W. R. Sherman and E. Robins, ibrd., 40, 803 (1968). 1044
ANALYTICAL CHEMISTRY, VOL. 44, NO. 6, M A Y 1972
Apparatus. Absorption spectra were obtained using a Beckman DB-GT spectrophotometer. Fluorescence measurements were performed on a Perkin-Elmer MPF-2A fluorescence spectrophotometer whose monochromators were calibrated against the xenon line emission spectrum and whose output was corrected for instrumental response by means of a rhodamine-B quantum counter. Reagents. Coumarin was purchased from Aldrich Chemical Co., Cedar Knolls, N. J., and 4-methyl-7-hydroxycoumarin was obtained from K and K Laboratories, Plainview, N.Y. Both compounds were recrystallized from 95 ethanol. The 4-methyl-7-methoxy derivative was prepared by methylating 4-methyl-7-hydroxycoumarin with dimethyl sulfate and potassium carbonate in acetone. The methylated product was
recrystallized from 95% ethanol and had a melting point range of 157-158 "C which agreed with that in the literature
Table I. Features of Absorption and Fluorescence Spectra of Coumarin and Some 7-Substituted Coumarin Derivatives in Strong Acid Long wavelength absorption maxima Fluorescence maxima
(1.2).
Analytical Reagent grade sulfuric acid was purchased from Mallinckrodt Chemical Works, St. Louis, Mo., and was used without further purification. Solutions of varying acidity for fluorimetric and absorptiometric titrations were prepared by dilution of the sulfuric acid with distilled deionized water. RESULTS
Fluorescence Spectra. At p H values above 2, the 4-methyl7-hydroxycoumarin has a peak emission at 450 nm. As the acidity increases, a bathochromic shift occurs and a less intense fluorescence band (peak emission 480 nm) develops. These data are in perfect agreement with the findings of Fink and Koehler in their studies with 7-hydroxycoumarin ( I O ) . These authors did not work in a p H range more acidic than 1.2. However, if the acidity is increased further, an intense blue fluorescence band appears with a peak emission at 412 nm. The 4-methyl-7-methoxycoumarin derivative does not show the same behavior as the 7-hydroxy compound. In fact, as the acidity increases, below pH 1, a bathochromic shift in fluorescence emission is observed. From pH 1-3, the fluorescence band remains constant in intensity with a peak emission at 386 nm. From H o = -4.7 to -10, the peak emission occurs at 412 nm. Coumarin, itself, has been reported to be nonfluorescent over a wide pH range ( I , 13, 14). We have found, however, that the protonated coumarin molecule is fluorescent and a titration curve for this molecule can be obtained in sulfuric acid solutions. The protonated coumarin has an emission maximum at 440 nm. The neutral molecule apparently has little or no fluorescence, although from H o values of -4.7 to -0.73, there is a gradual hypsochromic shift in emission toward 400 nm. The fluorescence intensity is so weak below H o = -3 that it is not possible to accurately determine the emission maximum of the neutral coumarin molecule. Absorption Spectra. The ultraviolet absorption spectra of all of the coumarins studied vary with acidity in the sulfuric acid media. Table I gives the pertinent spectral characteristics of the compounds in sulfuric acid solutions. A comparison of the dissociation constants obtained by absorption and fluorescence methods is given in Table 11.
Ho = 1 . 0 -8.92 A, nm A, nm
pH
Compound Coumarin 4-Methyl-7-hydroxycoumarin 4-Methyl-7-methoxycoumarin ~~
=
pH
=
1.0
Ho = -8.92
A, nm
A, nm
-400
440
312
322
322
345
476
41 2
322
352
386
412
~
Table 11. Dissociation Constants for Some Protonated Coumarins Determined by Fluorescence and Ultraviolet Spectral Titration Methods PKa
Compound Coumarin 4-Methyl-7-hydroxycoumarin 4-Methyl-7-methoxycoumarin
(fluorimetry) -5 3 -4 1 2.2 -4.3
P Ka (absorptiometry) -7.4 -5.0 7.8
-5 8
phenolic group is not consistent with literature reports that phenolic groups generally tend to become more acidic in the lowest excited singlet state relative to the ground state (15,16). The expected shift would therefore be from shorter to longer wavelength on dissociation. Second, the present study shows another prototropic reaction occurring at even higher acidities giving rise to an intense blue fluorescence with a shift in emission maximum that would not be consistent with protonation of the carbonyl group of the neutral coumarin molecule. It seems unlikely that the fluorescence of 4-methyl-7-hydroxycoumarin in 0.01M hydrochloric acid is due to the neutral molecule as was stated ( 3 ) ; nor does it seem likely that the shift in emission maxima in the pH range 1 to 3 is due to excited state dissociation of the neutral molecule as was postulated ( I O ) , although in this pH range only the neutral species exists in the ground state. The possible protolytic equilibria involving the addition of protons to the 4-methyl-7-hydroxycoumarin molecule can be described as in Scheme 1.
DISCUSSION
Chen (3) reported a 457-nm emission maximum for 4methyl-7-hydroxycoumarin in 0.01M sodium hydroxide and 474-nm emission maximum in 0.01M hydrochloric acid. The author assigned these maxima to the anionic and neutral forms of the molecule, respectively. The fluorescence spectra for 4methyl-7-hydroxycoumarin obtained in the present study are in agreement with those reported for the two specific p H conditions mentioned by Chen (3). The present data are also similar t o those reported for 7-hydroxycoumarin in the p H range 1.2 to 11.2 (IO). While the previous literature ( I O ) stated that this shift was the result of an excited state protolytic dissociation of the neutral coumarin molecule, two factors indicate that this may not be the case. First, the observed shift from a longer to shorter wavelength o n dissociation of a (12) M. Narayana, J. F. Dash, and P. D. Gardner, J . Org. Cliem., 27, 4704 (1962). (13) B. N. Mattoo, Trmis. Faraday Soc., 52, 1184 (1956'1. (14) C. E. Whelloch, J . Amer. Cl7em. Soc., 81, 1348 (1959).
11
.mt n
H'
m Scheme 1 In the neutral and alkaline pH range, the blue fluorescence of the excited 4-methyl-7-hydroxycoumarin molecule is due to (15) A . Weller, Progr. React. Kiuef., 1, 187 (1961). (16) W. Bartok, P. J. Lucchesi, and N. S. Snider, J . Amer. Clrrm. Soc., 84, 1842 (1962). ANALYTICAL CHEMISTRY, VOL. 44, NO. 6, MAY 1972
1045
the anion, I. As the acidity is increased, there are two possible sites of protonation-either at the dissociated phenolic oxygen, leading to the neutral molecule, 111, o r at the carbonyl oxygen, forming the zwitterionic species, 11. I n very strong acid media, the second protonation will lead to the cation, IV, whether I1 or I11 is the uncharged species from which it is derived. If the stepwise addition of the protons led to a path such as I 111 IV, one would expect to see the fluorescence bands shift from the blue of the anion, I, to shorter wavelengths o n protonation of the phenolate oxygen, and finally to longer wavelengths again, o n protonation of the carbonyl group to give IV. These are the changes observed in the absorption spectra as a function of acidity indicating that I11 is the predominant uncharged species in the ground state. The fluorescence spectral shifts, however, change from longer to shorter wavelengths as the acidity increases in concentrated sulfuric acid. Consideration of the pathway I -., I1 + IV indicates that this route in the excited state, will produce the required fluorescence spectral changes. That the blue-green fluorescence of 4-methyl-7-hydroxycoumarin does not originate from the excited neutral species (111), is supported by the acidity dependence of the emission spectra of the 7-methoxy derivative and that of coumarin itself. Neither of the latter compounds contains a dissociable proton and thus neither can form a n anion or zwitterion. In alkaline, neutral, and dilute acid solutions, the fluorescences of coumarin and the 7-methoxy derivative are unchanged, very weak, and occur at -400 nm and 386 nm, respectively. These emissions, especially that of the 7-methoxy derivative, arise from an electronic configuration very close to that of the excited neutral species. I n concentrated acid, both molecules are protonated, presumably at the carbonyl group, and the emission spectra as well as the absorption spectra shift to longer wavelengths (Table I) in accordance with the predicted emission behavior for protonation at a caibonyl group. I n chloroform, whose low dielectric strength and weak hydrogen bonding properties might not stabilize the excited zwitterion, the 7-hydroxy derivative fluoresces at 380 nm, presumably from the true neutral species. That the bluegreen fluorescence of the 7-hydroxy derivative does not originate from a dimer is supported by the lack of dependence of the emission wavelength and the linearity of the fluorescence signal, at p H 1, with varying 4-methyl-7-hydroxycoumarin concentration in the range 1 x 10-3M-l x 10-7M. Thus the zwitterion (11) appears to be the predominant uncharged species, derived from 7-hydroxy-4-methylcoumarin in the lowest excited singlet state, indicating that in the latter electronic state the carbonyl oxygen is more basic than the phenoxy anion. I n the range p H 2 t o Ho -4, the formation of the excited zwitterion must proceed entirely by tautomerization of the excited neutral species, since the latter is the sole absorbing species in this range. This requires a n extremely rapid twoproton transfer, to the solvent by the hydroxy group and from the solvent t o the carbonyl group. Presumably this process is mediated by hydrogen bonding with the solvent in the FranckCondon excited state of the neutral molecule and indicates that even moderately concentrated sulfuric acid-water solutions have substantial proton-accepting ability. A rather interesting application of this study lies in the interpretation of the fluorescence changes observed when
-
1046
ANALYTICAL CHEMISTRY, VOL. 44, NO. 6, MAY 1972
essentially nonfluorescent esters of umbelliferone and its derivatives are enzymatically hydrolyzed to yield fluorescent species derived from umbelliferone (4-8). I n the ester form, umbelliferone is covalently bound through the 7-hydroxy group to a substrate and is constrained to a n electronic configuration analogous to that of the neutral species or methoxy derivative. Thus the ester is weakly or not at all fluorescent, depending upon the nature of the interactions of the substrate with the umbelliferone moiety. Enzymic hydrolysis liberates the free umbelliferone moiety which rapidly equilibrates in the excited state, subsequent to excitation, to the excited anion in neutral solutions, resulting in intense blue fluorescence, o r to the excited zwitterion in acidic solutions, giving rise to blue green fluorescenced If the neutral species derived from umbelliferone did fluoresce intensely, studies of enzyme kinetics by this method might be seriously complicated by overlap of the fluorescence of the bound moiety with that of the free fluorescing species. The data of Table I1 show that the pK, values obtained by fluorimetry and absorptiometry in concentrated sulfuric acid differ. This would be expected if both excited state and ground state prototropic reactions are occurring in this acidity region. However, quantitative estimation of the dissociation constants for the ground state (and possibly excited state) protonations of the carbonyl groups of coumarin and the 7-methoxy derivative, and for the excited state protonation of the phenolic group of the 7-hydroxy derivative are complicated by the fact that the coumarins undergo some type of reaction in sulfuric acid solutions as evidenced by changes in the absorption spectra with time. The absorption spectra as a function of acidity did not show isosbestic points indicating that more than a simple transformation between acid and conjugate base was occurring. Thus, the pK, values obtained in strongly acidic sulfuric acid media should only be taken as approximate even though the measurements were made as quickly as possible. RECEIVED for review October 12, 1971. Accepted December 13, 1971. Taken in part from a dissertation submitted by 6 . J. Yakatan to the University of Florida, in partial fulfillment of the Doctor of Philosophy degree, December 1971.
Correction Simultaneous Automated Determination of Hydralazine Hydrochloride, Hydrochlorothiazide, and Reserpine in Single Tablet Formulations In this paper by Tibor Urbrinyi and Arthur O’Connell [ANAL.CHEM.,44, 565 (1972)], the authors would like to add the following acknowledgment. “The authors thank Henry Stober for his advice in using for separation the ion exchange resin.”
