T&mu, Vol. 36, No. 3, pp. 416-418, 1989 ~39-91~/89 $3.00 + 0.

00
printed in Great Britain. All rights reserved Copyright 0 1989 Pergamon Pres pie

STUDIES ON FLUORESCEIN-VII

THE FLUORESCENCE OF
FLUORESCEIN AS A FUNCTION OF pH

HARVEYDIEHL~ and RICHARDMARKUSZEWSKI$

~~~~rn~t of Chemistry and @rnes Laboratory, Iowa State University, Ames, IA 50011, U.S.A.

(Received 21 October 1987. Accepted 21 October 1988)

Summary-The relative fluorescence of fluorescein over the pH range 3-12 has been measured at 516 nm,
with excitation at 489 nm. The relative fluorescence is essentially zero at pH 3, increases slowly between
pH 4 and 5, rises rapidly between pH 6 and 7, reaches a maximum at pH 8, and remains constant at above
pH 8. The curve of relative fluorescence as a function of pH lies somewhat above the corresponding curve
describing the fraction of fluorescein present as the doubly charged anion, F12-, indicating much weaker
fluorescence of the singly charged anion, HFl -, and very much weaker fluorescence by the neutral species,
H,Fl. The fluorescence data have been used to calculate a value for the third dissociation constant. Because
of the complexity of the system, one unknown dissociation constant and three (relative) fluorescence
constants, a series of three variable regressions on the data was made. The final values were
K,,, = 4.36 x lo-’ (JI = 0.10) for the third dissociation constant and ~~~~ = 0.8; K”~, = 5.7; K~, = 100.0for
the relative fluorescence constants.

The primary interest in fluorescein lies in its
fluorescence, and it is therefore surprising that the
relation of this to the nature of the prototropic forms
of fluorescein in water solution has not been un-
equivocally established. Some workers have pro-
ceeded on the basis that the doubly charged anion
F12- is responsible for the fluorescence, inasmuch as
alkaline solution is necessary for the fluorescence to
appear, but this assumption has been made without
knowledge of the values of the acid dissociation
constants and the range of pH over which the various
prototropic forms exist and the fluorescence occurs.
An extensive study of the fluorescence was made by
the flash photolysis method by Lindqvist.’ Finding a
knowledge of the dissociation constants of fluorescein
necessary, he determined them by the spec-
trophotomet~c method, by the procedure described
by Zanker and Peter,2 but in purely aqueous solution.
Because of the difficulties posed by the closeness of
the second and third dissociation constants, a
difference of only 2.12 in pK, the Lindqvist value for
the third dissociation constant, KHFI = 1.99 x IO-’
(PKHFI= 6.7), is probably too small. Although
Lindqvist was concerned with the dissociation con-
stants of the excited states and the decay rates as a
function of pH, nowhere in the paper did he report
the observed relative fluorescence as a function of
PH.
We have now measured the fluorescence of highly
purified fluorescein in water as solvent, at ionic
strength 0.10, pH 3-12, with excitation at 489 nm and
*Part VI-H. Diehl, Talunru, 1989, 34, 413.
416
emission measurement at 516 nm (Table 1 and Fig.
1, curve F’).
The fluorescence is essentially zero at pH < 3,
increases slowly between 4 and 5, rises rapidly be-
tween pH 6 and 7, reaches a maximum at pH 8, and
remains constant at pH 2 8. Visual inspection shows
the point of inflection of the curve to be at pH 6.35.
Figure 1 also shows the distribution of the four
prototropic forms of yellow fluorescein, the only
form present in aqueous medium. The curve for the
fluorescence lies slightly above that for the fraction of
F12- present, especially at lower pH. It is apparent
that Fl- also fluoresces, but not as strongly as Fl*-,
and that quite possibly the undissociated species also
fluoresces but even more weakly than Fl-.
EXPERIMRNTAL
Chemicals
Fluorescein was prepared and purified through the
diacetyl compound described in an earlier paper.)
Buffers covering the pH region 1.5-13 at intervals of 0.5
were prepared from 0,IM hydrochloric acid, potassium
hydrogen phthalate, boric acid and potassium hydroxide, as
appropriate, with addition of 0.W potassium chloride to
maintain the ionic strength constant at p = 0.1. After the
fluorescence measurement the pH of the solution was deter-
mined with a Corning Model 10 pH-meter and a Beckman
glass electrode and SCE.
Standards
A stock solution of fluorescein was prepared by dissolving
100.0 mg of pure red fluorescein in 1 litre of 0. 1M potassium
hydroxide. A 1.00~ml afiquot of the stock solution was
placed in a SO-ml standard flask, 1 ml of 0.1M hydrochloric
acid was added, and the mixture was diluted to the mark
with the appropriate buffer. The concentration of
fluorescein in all the solutions tested was 6 x 10m6M.
 ANALYTICAL DATA 417

Table 1. Relative fluorescence of fluorescein as a function of pH [Pm= observed relative fluorescence

normal&d to maximum fluorescence @H > 8) set equal to 100.0, ionic strength p = 0.10; concentration
6.0 x 10-6M]
;H 0.92
3.11 3.56
1.67 4.05
3.00 4.16
4.34 4.65
5.80 4.84
7.10 10.0
5.11 11.6
5.28
m
PH 5.35 5.50 5.51 5.69 5.85 6.05 6.35 6.63
Fb 13.5 17.4 16.7 21.7 27.1 38.2 52.5 67.5
PH 6.99 7.48 7.73 8.20 8.40 9.10 10.45 12.40
F:, 82.7 95.8 99.2 100.0 100.0 100.0 100.0 100.0

Procedure The individual fluorescence intensities are given by

The excitation and emission spectra of fluorescein as a
function of pH were obtained by use of a Bowman-Kiers FHzFl= 2.303 4H2Fl %aln b[H,nl
Spectrophosphorimeter with an attached Mosely Autograf
X-Y Recorder. The path-length of the fused-silica cell was = ~,,[H,Fll (2)
1.00cm. The emission at 516 nm was recorded over the pH
&I = 2.303 &F, &HFl b WI- 1
ranae 3.1 I-12.4, with excitation at 489 nm.
= ~HFI[HFI-I (3)
RESULTS AND DlSCUSSlON FF,= 2.303 & +, b[F12-]

For the prototropic forms and dissociation _ con- = k,[F12-] (4)

_.
stants of fluorescein, we employ the same symbolism
where 4 is the fluorescence efficiency, E is the molar
as in the earlier papers. 3-7 For definition of KH,m,
absorptivity, b the path-length, and k the fluorescence
KH,, and 4~ 9 see reference 4. For calculation of the
constant. In equations (2)-(4) the fluorescence
distribution of the four prototropic forms (H3Fl+,
efficiency and the molar absorptivity for a given form
H,Fl, HFl- , Fl*-) see reference 7.
are assumed to be constant and the path-length to be
We now assume that the fluorescence intensities, F,,
1.00 cm. Combination of equations (l)-(4) yields
of the individual prototropic forms are additive to
give the total fluorescence F,,,: Fm= kH,d-hFll+ k,,,[HFl- I+ kmW2 - 1 (5)
Fn,= FH,, + FHF,+ 4-1 (1) The concentration terms are eliminated by substi-

100

0
3.0 5.0 7.0 9.0
PH
Fig. 1. Normalized relative fluorescence F:. of 6.0 x 10-6M fluorescein as a function of pH; excitation
at 489 nm, emission at 516 nm, ionic strength p = 0.10. A, B, C and D are the fractions of fluorescein
present as H,Fl+, H,Fl, HFl- and Fl*- respectively as calculated from p&r, = 2.19, pKHzF,= 4.24,
pK,, = 6.36.
418 ANALYTICAL DATA

tution of the equations defining the concentrations of optimum fit of observed and calculated fluorescences
the individual prototropic forms as fractions (a,) of is obtained.
the total concentration C: The values assumed for the first and second dis-
sociation constants, pKu,, = 2.19 and PKH,~ = 4.24,
aH*F,=y obtained from the ultraviolet absorption studies re-
ported earlier,6*7 were held constant throughout the
QHF, --WF1
- -1 (7)
calculation. The third dissociation constant was var-
C ied in a series of regressions: pKHF,= 6.330, 6.350,
6.355, 6.360, 6.365, 6.370, 6.390. As a criterion of fit,
ct, - [FIZ-) the differences between the observed and calculated
(8)
c
values for the total fluorescence were calculated; the
giving standard deviations and sums-of-squares showed
pronounced minima at pKHn = 6.360, at which the
% = kn,n aH2FI c i- kH, QHFI c + kn an C (9)
standard deviation of an individual observation
the prime indicating that the values of Fi have been (n = 22) was 0.694 and the standard deviation of the
normalized. With C constant throughout a series of mean was 0.15 1 (both in terms of relative fluorescence
measurements, a new term, K, the “working normalized to 100.0). At this point i$“= 0.804,
fluorescence constant”, where K = kc, is introduced ~~~~= 5.819, and rcr, = 101.45.
for each of the prototropic forms, giving We conclude that the third dissociation constant
KHFlIS equal to 4.36 x lo-’ (PKH, = 6.36) and nor-
F6 = KH2Fl aH2fl + KHFl OLHR + 61 aFi (10)
malizing to icF1= 100.0 gives the working fluorescence
It is apparent that the system involves six con- constants as Knlri = 0.8, xHF,= 5.7, and K~ = 100.0.
stants: the three dissociation constants, KH2n, KH2m,
K HF,r and the three working fluorescence constants,
KH2R, lCHFl and K~. Our present approach to this REFERENCES
problem has been to assume values for the three 1. L. Lindqvist, Arkiu Kemi, 1960, 16, 79.
dissociation constants, calculate the fractions of the 2. V. Zanker and W. Peter, Ber., 1958, 91, 572.
various prototropic forms present, as functions of 3. R. Markuszewski and H. Diehl, Takmta, 1980,27,937.
pH, substitute these in equation (10) and apply 4. H. Diehl and R. Markusmvski, ibid., 1985, 32, 159.
5. H. Diehl, N. Horchak-Morris, A. J. Hefley, L. F.
regression to evaluate the three working fluorescence
Munson and R. Markuszewski, ibid., 1986, 33, 901.
constants, and then to repeat the process with a 6. H. Diehl and N. Horchak-Morris, ibid., 1987, 34, 739.
slightly different set of dissociation constants until an 7. H. Diehl, ibid., 1989, 36, 413.