84 SHORT COi%tBlUNICATIONS

The transition potentials of the three indicators are + 0.17 V, + o.rG V and + 0.10 V for safraninr! T, phenosafranine and neutral red rcspcctivclyn; these values corre- spond to the potentials at which the colour is discharged in 1.0 N hydrochloric acid solutions containing uxalic acid under carbon dioxide at 28*.

The indicators reported have the following advantages over methylenc blue. Under the conditions prescribed, it is not necessary to wait for 2-3 set after each addition of titrant near the en&point. Methylenc blue indicator must be added in quantities just sufficient to colour the solution blue, because of the large indicator correction The dyes recommenclccl above have much better tinctorial properties so that a brilliant colour can be obtained without an indicator correction being needed, Finally, the colour changes arc cluite sharp even at room temperature, whereas mcthylenc blue requires a temperature of 50-70”.

One of us (P.V.IC.l<.) thanks the Council of Scientific and Industrial IZeseurch, India, for the award of a Junior Research Fellowship.

(Received November sth, rgOz)

Complex formation and fluorescence Part 111. Salicylate complexes

ft is possible to detcrminc salicylates in low concentration by the measurcmcnt of fluorescence intensity ia”. This is usually carried out between pH 0 and 8 using the fluorescence of the singly charged snlicylate anion. There have been only three investi- gations of the effect of the presence of cations on the intensity of this fluoresccncc. GOTO~ has reported that iron(II1) causes complete extinction of fluorescence. ZITTI.E et al, and also ROSEN AND WILLIABIS~ have shown that at a PH of IO, the presence of magnesiutn(rK) causes a shift in the peak of the fluorescence emission spectrum from that due to the singly charged salicylate anion (410 mp) to that of the doubly charged anion (390 mp), with an accompanying increase in intensity of fluorescence.

The writer has investigated the effect of adding cations to salicylate solutions over the pn range 2 to 7. The cations were chosen in order to study several variables: ion size, ion charge and the difference between transition metals and non-transition

SHORT COMMUNICATlONS 85

metals. Ions used were Cu(II), Co(II), Ni(II), Zn(1.I). M&II), Re(IL), Fe(II1) and -ku(Ur).

Solutions for fluorescence measurenlents of bivalent cations were prepared by dilution of solutions in which the ratio Of cation to salicylate was I to 2. In the case 0f Fe(rrx) and Aqrrr) solutions the effect of a small concentration relative to that of snlicylate was desired, since GOTO 3 hnd rcportcd that very small amounts of FC- (III) would completely destroy the salicylate fluorescence. Al(III’) was chosen for comparison with Fe(III). Salts used were of reagent grade, and the salicylic acid was obtained from the Eastman Kodak Co. The PH was adjusted before dilution of the solutions, using dilute hydrochloric acid or sodium hydroxide, and waschecked after dilution using a Reckman pH meter. All fluorcscencc mcasuremcnts wcrc made on the Farrand spectrofluorimcter previously described 0, The excitation wavelength was set at 6x0 rnp and a complctc emission spectrum was dcterminccl for each solution between 330 rnp ancl 550 m,u.

I?ISCUSSION OF RESULTS

Of the ions st.udiccl only Cu(lI), mqrr), n(m) and iil(rrr) caused any change in the ffuoresccncc, either with regard to intensity of fluorescence or range of emission spectrum. Results for Cu(I1) and He(lI) arc sl~wn in Fig. I. The effect of small amounts of r+(m) and Al(m) on the fluorescence intensity of salicylatc solutions is shown in Fig. 2. ‘The solutions containing l~c(III) chanfgzd on standing. The results in Pig. 2 were obtained immcdiatcly after the solutions were prepared. Go-r03 found that the fluorescence of salicylate solutions was clcstroyed by 2.5 pg Of Fc(III) with no mention being given of relative amounts or ctxuxxttrations.

As can be seen from Fig. x, the presence of Cu(II) produced a decrease in fluores-

Fig. I. Effect of pH, ac(II) nnd Cu(II) on sdicylatc fluorcsccncc intensity. (a) 13~ -t_ I~Sal- (380 mp); (b) WSal- (400 mp): (c) (Cu + WSal- (400 mp): (d) I-Ed- (380 mp).

Apia/. C&h. Rcln, 29 (14tj3) 84-87

86 SHORT CoIfMUNICATIONS

cencc intensity of salicylatc solutions, the decrease becoming greater as the PH

incrcasecl. ‘This is of particular interest in biological analyses, since SCHUUEIZT~ has stated that there is some connection bctwecn the physiological action of salicylates and the presence of copper.

In the case of 13e(II), there was not only a pronounced increase in fluorescence intensity, but also a shift of the emission spectrum so that the peak occurred at 380 mj2, shifting from the 500 rnp of the sodium salicylatc solutions. (It should be noted that the ert&sion curve is very flat at the “peak”, but the shift is quite definite.) This shift did not take place with other solutions. A similar shift rcportcd for solutions of sodium salicylate containirq Mg(II) at a pH of 10 de&, has been explained as being due to the formation of a complex ion in which both llyc~ro~cns of the salicylatc have been displaced, producing the fluoresccncc spectrum and intensity to be cxpectod for the doubly charged salicylntc anion. ‘This anion does not form in sodium salicylate solutions below a pH of x410”. l’hc intensity of the fluorcscencc of the 13c(H) complexes was approsimately 1.7 times that of the snlicylatc anion at n plr of 7 measured at 400 mp, arid almost 4 times that of salicylrttc ion at p,r 7 at 380 rnp,

f-~f~XTNEI+-~lZC3JN ANI) COHI;NH concluded from investigationsof the Cu(II)-snlicylntc system by spectrophotomctric mcthocls that it is not possible to dcciclc whether the ML:! complex is one in which the phenolic hydrogen has been displnccd, and suggested that tlicrc may be a mixture present. ‘I’hc beryllium complex rcportccl by ScHuuorzr ANI) LINL~ENWU~!” is an MHL type, in which the phcnolic hydrogen has not been displaced. ‘The emission spectra obtnincd in the present work indicate that the beryl- lium complex is of the ML2 type in which both hydrogcns of the salicylic acid hnvc been clisplncccl. Since the copper complex clots not fluorcscc, no direct cvidcncc is obtained with regard to tltc problen~ set out by HEITNISR-IRCUN AND COHEN. It may be possible to obtain the answer to this qucstiou indirectly by fluorescent studies, and work is being carried on with this in mind. CO%C&tSiO?tS

Only cations which form complexes with salicylatcs I ‘10 have any effect on the fiuores- cence. The results parallel those of STEVENSON wit0 fbund that ions of the transition elements did not cause fluorescence in S-quinolinol complcxcs. The effect of a very small ion which forms strong oxygen bonds is shown in the case of. I3e(II), which

SHORT COAfMUNICATIONS 87

produced both an increase in fluorescence intensity and also a change in the peak wavelength of the emission spectrum.

This investigation was supportccl by research grant No. 8005, from the National Institutes of Health, U. S. Public Health Service,

Newzvlt College of Engineeriq. Newark 2, New Jersey (U.S.A.)

JOHN A. BISHOP

(Kcccivud f;clxuary zznd, 1963)

A sensitive method for end-point detection in constant current coulometry

Constant current coulometric titrations of acicls and bases have been described carlicrl. For high precision analysis with such a method the exact determination of the endpoint is important. I’otcntiometric methods of indication are very often used and we have therefore examined the sensitive detection method described clsewhcrc2, for acid-base coulometric titrations. In the ~trai~l~tfor~vard ~otentiomctric titration, the endpoint is locntcd by recording in the close ncighbourhoocl of the endpoint the derivative of successive potential changes resulting from successive additions of very smail quantities of titrating reagents (e.g. 0.010 ml of 0.1 N sodium hydroxide solution).

During a coulometric titration the equivalent of such successive additions can be realized by passing a current of exactly known intensity during a short time interval, e.g. 50 or xoo mA during I set, xoo mA for I see being about equivalent to the above mentioned amount of base. A similar technique has been described for a coulometric determination of phenol3. These short impulses of constant current intensity are produced by using a preset-time timer in connection with the current generator. The coulometer is calibrated by measuring the potential drop across high precision resistances that are included in the eIcctrolysis circuit. Two resistances IRt and R2,

Fig. I) ore used to control the current during the electrolysis and are themselves