Volume 121. number 6

CHEMICAL PHYSICS LETTERS

EXCITED-STATE

PROTON

Anunay SAMANTA, Tapanendu KUNDU

TRANSFER

KINETICS

NItIn CHATTOPADHYAY. and Mhir CHOWDHURY

22 November

1985

OF CARBAZOLE

Debnarayan

NATH,

Depa~~menr of PhyssrcolChemrrty Indmn Assorrafron for the Cul~rcorron of Scrence. Jadoupur. Calcurra 700 032 Indro

Rgflved 30

July 1985 II-I final form 6 September 1985

Rate constenLs for the excaed-sue pro1011transfer reaction ol carbazole m aqueous alhahne soluuon have been delermrncd usmg pncosecond single photon counring Fluorescence decay measurements show Ihat the bath renctmn 15slow compared 10 the fluorescence decay ume and theretore equrhbnum IS not attamed m the exerted slate The valrdrty of a pK value Ior ths lowest excited slate deiermmed from aeady-slate fluorescence measuremenls ;rrsummg eqnbbnum 1s dlscwsed. 111s concluded that the thermodynamx pK’ l&e for carbazole IS 10 98

L Introduction In many cases the chemrcal and physical propertres of electronically exerted molecules differ markedly from those of ground-state molecules because of the different electronic distribution m the exerted state [l-7]. The ionization constants of electronically exerted

orgamc

of magnitude

acrds and bases differ by several orders from

those

observed

m the ground

state

[3-71 The methods of determination of pK* are mainly based on steady-state absorptron and fluorescence spectra Most studies make use of the Forster cycle [3.4] in which the difference between the absorptron or emusion spectral ongm of the acidic form and its conJugate base IS attnbuted to the difference in pA’values between ground and excited states Steadystate fluorescence intensity techniques developed by Weller [4,5] are also used to determme pK* values, m which relative quantum yields of the excited-state species are plotted agamst the pH of the solutron. The former method, althoughvery successfulinmterpreting acid-base

properties

of several

organic

compounds,

some questionable assumptions [8,9] and it is now concluded that the Fbrster cycle gives precrse pK* values only in some special cases [ 6]_ The latter method suffers from the disadvantage that rt requires prototropic equilibrium to be established in the excitmvolves

0 009-2614/85/S (North-Holland

03.30 0 Blsevier Science Publishers B V. Physrcs Pubhslung Drvision)

ed state, and makes some assumptions regarding the fluorescence hfetrme of the exerted-state specres [6] However, the sunple question as to whether equilibnum is attained

at all in the excited

state

cannot

be an-

swered by either method It 1s Interesting to see whether the pK* determrned from steady-state lluo rescence measurements represents a “true” pK*. It is unhkely

that

equihbrium

will be attamed

when

only

one of the two excited-state species IS fluorescent, because then the non-emittmg species has probably a hfetime of the order of lo-L1 s. l&ser and Feitelson [IO], on the basis of this argument, concluded that m such cases the measured pK*, obtained from the pH dependence

of only

“true”

pK*

The actual

higher

or lower

than

one form,

did not

represent

thermodynamic

the “apparent”

pK*

the

may be

pK*

Smce the question of attainment ofequ~librium can be solved by direct kmetlc measurements of flue rescence decay oi excited-state species, “true” pK* values can only be obtamed

from

such measurements

Recent kinetic measurements [l I-181 have proved the superionty of tius approach over steady-state methods. Direct determination of rate parameters can provide informatron unobtamable from othermethods [12,141 In order to gam insight into the proton transfer mechanism, we have studred the excited-state proton 507

Volume 121.

number 6

22 November 1985

CHEMICAL PHYSICS LETTERS

transfer kmetrcs of an N-heterocychc compound, Carbazole (Aldrich) was prified by repeated crysin the excited state, and has emitting neutral and anionic forms

by inspection of reduced chi-square plot of weighted residuals and autocorrelation functions of the residuals) was recorded by a plotter

3 Results and drscussron 2. Experimental Carbazole (Aldrrch) was purrfied by repeated crystallisatton from 75% alcohol The recrystallised sample was vacuum sublimed, column chromatographed using a neutral alumina column and 20% benzenepetroleum ether as eluent, then recrystallrsed once more Its punty was checked by meltmg point determmatron and spectroscopic measurements Tnply distrlled water was used for prepanng solutrons Sodrum hydroxide (E Merck, A R) was used without further purrficatron.pHvaluesofthesolutionswere measured by standard analytrcal methods The solutrons were not degassed as this does not change the hfetirne of carbazole Absorptron and fluorescence spectra were recorded by a Car-y 17D spectrophotometer and Perkin-Elmer MPF 44B spectrofluonmerer, respectrvely The errus ston spectra were not corrected for Instrumental response For fluorescence decay measurements, the flash lamp of nanosecond smgle-photon counting spectrofluonrneter (Apphed Photophysrcs. SP70) was replaccd by a mode-locked argon-ion pumped dye laser (Spectra Physics) for greater accuracy. The laser, operated at 800 kHz repetitron rate, produced reproducible trams of pulses of fwhm = 30 ps The cavity dumped dye laser output was frequency doubled and used as the excitation source (295 nm) A grating monochromator (model 7300, Applied Photophysics wtth aperture ratlof/3.4) was used in the emission side whereas two visrble cut-off filters were used in the excitation side to block the direct dye laser beam A multrchannel pulse height analyser was used to obtam the decay curve of the sample and the lamp profile The sample decay profile is expected to be distorted by the Brute time response of the detection system Computer analyscd deconvolutron was, therefor, necessary to obtain the decay functron. The decay curves from multrchannel analyser were transferred to a Sums 1 computer and deconvoluted The computer-analysed best fit (checked each time 508

The vanation of carbazole lununescence wrth pH IS shown in fig I_ At pH = 7, the fluorescence spectrum consists of two closely spaced peaks with maxima at around 360 nm. As thepWofthe solution is gradually increased, the mtensrty of the 360 nm band decreases and a new rnaxrmum appears at 415 nm. At this pH range the absorption and excrtation spectra, however, remam unchanged mdrcating the occurrence of an excited-state process The 415 nm band, as described by Schulman et al [ 19 ] is due to enussion from carbazole anion produced m the excited state From fluorescence titration they obtamed a value of 11.9 for the excited-state pK They also calculated indrrectly the unknown pK value for the ground state from the pK*, and also the shift in the fluorescence maxima, using a

s

vl ?!?

E w

320

340

360

380

Km

Wnvelenglh(ln

A20

nml

-

AA0

460

A00

rig 1 Emission spectra of cubazole as a function of pH The diffccnt alkah concentrations m spectra a + fare 0.0.0 005. 0 01, 0.02. 0 04 and 0 06 M respectively. The excitation wavelength is 295 nm

Volume

121, number 6

CHEMICAL

>.

I.

1

*

250 CHANNEL

1

;50

NUMBER

-

,

,

PHYSICS

I

PH

mncentration 000 0 005 001 0.02 0.04

0 06

C’.’

I::‘:

:: :; .::n ,’

L

I I

I

150 C-IANNEL

250 NUMBER

350 -

A50

Tlg 3 Fhoresccnoe decay CU~QX momtored nt 415 nm (Ac,, = 295 nm)- (-_) [OH-] = 0 005. (---) [OH-] = 0 01, ( --) [OH-] = 0 01, (- --) [OH-] = 0 04, (---I [OH-] = 0 06

Forster cycle The ground-state pK value thus obtamed IS21 1 The fluorescence decay curves for 360 nm and 415 nm at different pH are shown in figs 2 and 3 respectrvely. The decay of 360 nm band IS found to be single exponential and dependent on the pH of the solutron As the pH of the solutron IS mcreased the decay becomes faster The fluorescence decay curves analyzed at 415 nm show a build-up followed by a decay and can be resolved into a difference of two

Alkab

1985

\

-

decay pammeters of cubazole

22 November

450

Fig 2 Fluorescence decay curves momtored at 360 nm (XeXc = 295 nm) (-_) [OH] = 0 0, (- -) [OH-] = 0 005, ( ) [OH-] = 0.01, (-) [OH-] = 0 02, (---) [OH-] = 0 04, (-x -) [OH-] = 0 06, (---) pump profile

Table 1 Fluorescent

LETlXRS

exponential terms wtth equal pre-exponential factors Wrth increase of the pH of the solutron the build-up trmc is shortened. the follow-up decay, however, 1s found to be unchanged Table 1 summarises the hfetimes of the two components A kinetic scheme for excited-state proton transfer of carbazole rn an alkaline medrum 1s

at different pH

(N) (from 360

71

70 117 12.0 12 3

9 97 5.7 4.6 31

12 6 12 8

20 1.5

52

nm)

ON

Mean 72 (I-IS)

(from 415 run)

19.8 20 6 20 2 21.7

20.7

21.5 509

CHEMICAL

Volume 121. number 6

AH*

L -11ky+X_h

X-IW-‘-1=X-i A-*

X-&d 11 +

OH-

AH+OH-.

k2

x-10

PHYSICS LETTERS

A-

+ Hz0

x-20

[AH*]

where AH and A- represent carbazole and the correspondmg anion, k; and k2 are the pseudo-fist-order rate constants for the forward and the backward reactron in the excited state, k10 and k20 are the corresponding rate constants in the ground state, k, and k; are t!re rate constants for fluorescence of carbazole and amon respectively, kd and kh are the respective non-radratrve constants The drfferential equations that descnbe the above kmetrc scheme are

-d [A-*]

/dr = (k2 + kh + k;)[A-*]

where k-i = k, [OH-] Usmg the boundary conditrons and [A-*] = 0 at t = 0 we obtam

- ki [AH*]

, (2)

X1 A2 = (kf + k,)Y

iqs

e-X2f]

,

(e-h2r-e-Alf),

whereX=kf+kd

+kf+kl[OH-]

+ Y, _

(7)

to + (kh + k;)k, [OH- ]

(6) and (7) can be combined

A, +X2=kd+kf

(6)

(8)

to grve

+ [I - (kd + kf)/(kh +

+ k;)-h,&

@I Y (9

(3)

[AH*10

= Al

=k,

X,A2 =XY-kk,k2[OH-]

+ (k;

k, [OH-] [A-*](r)

A, +h2

Eq (7) is equivalent

= [AH*]~

exp [(kf + k,) t] _

Therefore, 111pure water carbazole will decay with a lifekf+ kd found from lifetime measuretime (kr+ k&l ments is 1.0 X lOa s-l The other rate constants can be evaluated easrly rf we can write eq (5) m the followmg manner

, (1)

bH*lo [AH*1 (f) = x, _ x2 X [(X - 1,) edAIr +(X1 -X)

= [AH*10

- k, [A-*]

[AH*]

1985

mined as follows. At pH = 7, the rate of proton transfer is negltgrble due to small [OH-], I e. kl [OH-] -=Zk, + kd and consequently [A-*] is very small so that k2 [A-*] = 0. Eq (1) reduces to -d [AH*] /dt = (kf + k,) [AH*] or,

+ H,.,

-d [AH*] /dt = (k; + k, + k,) [AH*]

22 November

-

+k;,

A2 Y=k;+kh

(4)

+k2> and

Xl, X2 = Til, 7F1 (5) Accordmg to thrs scheme, the emrssion of neutral carbazole should exhibit a double exponential decay with a ratro of long to short components (X, -X)/(XX2) The experrmental results, however, indicate a single cxponentral decay. Thrs deviatron from the proposed mechamsm IS more apparent than real, rt is caused by the very low value of the above ratio of pre-exponentral factors The decay curves obtained at 360 nm when analysed for biexponential decay grve two hfetunes, but wrth a much lower pre-exponential factor of the long-lived component_ The rate constants m the exerted state are deter510

OL

002

0 04

006

[OH-]-

Fy 4. (a) Plot of A, + A2 versus [OH-] for carbazole, the cloncentratmnsare expressed in molar umts (b) Plot of Al + hz versus ADA:! (c) Plot of hlh2 versus [OH-] for carbazole, the concentratmns are expressed in molar umtb

Volume 121. number 6

Table. 2 Rate constants and pK* of mbazole derived from fluorescence decay parameters Values 10 x 108 s-1 4.5 x 10’ s-1 9.0 x log hl-’ 85X 106s~’ 10 98

s-l

From the slope of the plot of h, + X2 against [OH-] (see fig 4a) we obtam a value of 9 0 X log M-l s-l for kI_ Accordmg to eq. (9) a plot of hI + h2 versus XIh2 would yield a straight line (fig 4b), from the slope of which k; + kb IS found to be 4.5 X lo7 s-I_ When XIX2 is plotted agamst [OH-] (eq. (8)) it grves a stra&t hne (fig 4~). The intercepts of figs 4c and 4a are utllised to evaluate k,, wluch comes out to be 8 5 X lo6 s-l The rate constants, found from this srudy, are summarised in table 2. The excited-state dissociation constants can now be calculated using the relation [13,15] log(k2/kI)

= pK* + log K,

,

Bves pK* = 10 98, which is lower than the value of I 1.9 obtained by us and others [19] from steady-state fluorescence measurements The dlscrep ancy between these two methods can be understood as follows. FollowmgWeller’s treatment [5] of the dependence of fluorescence mtenslties,

+ k;TO + k2Tb),

(11)

where 7. = (kf + kd)-l, ~b = (k; + kh)-l, and @/@o is the fluorescence quantum yield of carbazole relative to its maximum quantum yield Eq. (11) can be written as @//9o = l/(1

+R

X lo@),

where R = klrOK,/(l

+ k24

attained, but the inflection point would correspond to pK* only when equilibrium 1s reached and when the hfetlmes of the conjugate acid-base pan are Identical_ The present kinetic measurements, however, suggest that III this partrcular case equihbnum IS not reached; if there were an equllibnum then at pff= pK*, k; would have been equal to k2_ Actually, however, atpH= 11.9 @K*), k; = 7.1 X IO7 s-l whereas k2 (8.5 X IO6 s-l) is an order of magnitude less. It is also obvious that the reverse protonation process 1s much slower than the emlsslon decay process Reasons for the slowness of the protonstion and deprotonation process are bemg investigated further by varying the temperature, adding quenchers, studymg substrtuted denvatlves and by comparmg with an analogous nonplanar molecule, namely drphenylamme.

Acknowledgement The work has been project

(No.

Bhattacharyya slons

carned

23(lP-2)/81-STP

out under II)

the DST

We thank

Dr

K.

and Dr D S Ray for helpful dncus-

(10)

wluch

G/o0 = (1 + k27b)l(l

22 November 1985

CHEMICAL PHYSICS LEl-l-ERS

-

From tlus expression it is apparent that the relative fluorescence intensity versus pH plot would yield a slgmold curve u-respective of whether equhbrium 1s

References [I] J.B Birks, Photophysbs

of aromatic molecules (w~lcy, New York, 1970). [2] D-S Ray, K Bhattacharyya, S C Bera and M Chowdhury, Chem. Phys Letters 69 (1980) 134 [3] T. rdrster, 2. Llektrochem 54 (1950) 42. 531. [4] A Weller, Z. Elektrochem 56 (1952) 662.61 (1957)

956 [S] A Weller, Progr React Kmetics 1 (196i)

[6] J F Ireland and P AH

189. Wyatt. Advan Phys Org. Chem

12 (1976) 131. [7] E VanderDonct, Rogr. React Kmetics 5 (1970)

273 [8] T-C Werner and D hl. Hercules, J Phys. Chem 73 (1969; 2005. [9] H H Jaffcand HI- Jones, J Org Chem 30 (1965) 964 [lo] N Lasserand J Fextelson, J. Phys Chem 77 (1973) 1011 [ll] hi R Loken, J W Hayes, J R Gohlkeand L Brand, Biochemistry 11 (1972) 4779 [12] M Ofran and J. reitebon, Chem Phys Letters 19 (1973) 427 1131 A Gafm and L Brand. Chem Phys Letters 58 (1978) 346 [14] A Gafbi, RL Modlm and L Brand, J. Phys Chem 80 (1976) 898.

511

Volume 121, number 6

CHEMICAL

[15] J Shah, H C Pant and D D. Pant, Chem. Phys Letters 115 (1985) 192 [16] C_J Marzzacco. G Deckey and A M Halpern_ J Phys. Chem 86 (1982) 4937 [17] AJ. Camplllo. J H Clark, S L Shapiro and K R Wmn. m: Picosecond phenomena, Springer Scnes in Chemid Physxs. VoL4, cds C V Shank, E P Ippen and S L Shapiro (Spnnger, Berlin. 1978) pp 319 ff

512

PHYSICS LElTERS

22 November 1985

KB fisenthal, K Gna&g. Wm. Hethenngton. M Crawford and R Mxheels, in Piwsecond phenomena, Springer series m Chemical Physics, VoL 4, eds C-V_ Shank, E P Ippen and S L Shapiro (Springer, Berhn, 1978) pp 34 ff [19] A C Capomacctia and S G Schulman, AnaL Chim Acta 59 (1972) 471.

[18]