Kinetic isotope effect on the proton-transfer in indigo carmine

(1)

Kinetic isotope effect on the proton-transfer in indigo carmine

Izumi Iwakura

a,b

, Atsushi Yabushita

c

, Takayoshi Kobayashi

b,c,d,e,*

a

JSPS Research Fellow, 8 Ichibancho, Chiyoda-ku, Tokyo 102-8472, Japan

b

Department of Applied Physics and Chemistry and Institute for Laser Science, University of Electro-Communications, 1-5-1 Chofugaoka, Chofu, Tokyo 182-8585, Japan

c

Department of Electrophysics, National Chiao-Tung University, Hsinchu 300, Taiwan

d

ICORP, JST, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan

e

Institute of Laser Engineering, Osaka University, 2-6 Yamada-oka, Suita, Osaka 565-0971, Japan

a r t i c l e

i n f o

Article history: Received 7 October 2009 In ﬁnal form 13 November 2009 Available online 18 November 2009

a b s t r a c t

We have found that photo-excited indigo carmine causes stepwise proton transfer to generate mono-alcohol intermediate, which immediately returns to the parent molecule [I. Iwakura, A. Yabushita, T. Kobayashi, Chem. Lett. 38 (2009) 1020]. The kinetic isotope effect in the proton transfer of this system has been studied by direct observation of the real-time molecular vibrational amplitude during the reac-tion including the transireac-tion state using a sub-5-fs laser pulse. The kHa=k

D

aratio for the proton transfer after photo-excitation is observed to be 1.5 ± 0.1.

1. Introduction

Kinetic isotope effects (KIEs), usually deﬁned by the ratio of reaction rates between compounds with and without an isotopic label, have been used to clarify the mechanisms of many reactions, such as proton transfer[1–4], Claisen rearrangement[5–10], Diels– Alder reaction[11,12], and osmium oxidation[13].

However, it is difﬁcult to accurately estimate the reaction rates. The method of pump–probe measurement is based on the time-dependent molecular vibrations including their frequencies and phases, which directly provide information on structural changes. Therefore, it presents a crucial clue for clariﬁcation of the reaction mechanism behind the observed KIE.

In the present work, pump–probe signals have been obtained using a sub-5-fs laser pulse to observe real-time molecular vibra-tional frequency, which has enabled observation of the structural changes during the chemical reactions for direct estimation of KIE. We have chosen indigo carmine for our sample. We conﬁrmed in our previous work[14]that the stepwise proton transfer took place in this molecule after photo-excitation and that a mono-alcohol intermediate was generated. However, this intermediate was found to revert immediately to the parent molecule because this mono-alcohol species is unstable (Fig. 1). The present sub-5-fs study has enabled us to work on the KIE by this proton transfer without accumulation of the product.

2. Experimental

Indigo carmine in saturated anhydrous methanol solution (pur-ity >99.8%, <0.005% water, Sigma Aldrich) and methanol-d solution (>99.5 atom% D, Sigma Aldrich) (Fig. 2a) were selected as samples to identify the KIE in the proton transfer process in the excited state using the real-time vibrational spectroscopy, because indigo carmine has high enough solubilities in these solvents and metha-nol has no molecular vibrational signal in the frequency range examined in this study. A cell with an optical path length of 1 mm was used to contain the sample solution.

Pump–probe experiment using sub-5-fs laser pulses (Fig. 2b)

[15], which extended from 525 to 725 nm, was performed to iden-tify the KIE. Sub-5-fs pulses were generated by compressing the output from a non-collinear optical parametric amplifier (NOPA) seeded by a white-light continuum. The output of the sub-5-fs pulsed beam was split into two beams for pump-and-probe pulses. The pump source of the NOPA was a regenerative amplifier (Spec-tra-Physics, model Spitfire) with pulse duration of 100 fs, central wavelength of 790 nm, and repetition rate of 5 kHz.

Fig. 2c shows the intensity autocorrelation trace of the pulse after compression by using a BK7-prism pair with a 37 cm slant length between the two prisms. A ﬁt assuming a Gaussian function yielded a pulse width of 7.25 fs. The experiments were performed at pump and probe pulse intensities of 2580 and 480 GW cm2_,

respectively. The focal areas of the pump and probe laser pulses were 100 and 75

l

m2_{, respectively. The polarizations of the pump}

and probe pulses were parallel to each other. All measurements were performed at 295 ± 1 K.

The time-resolved difference transmittanceDT in the spectral range extending from 525 to 725 nm was simultaneously measured using a multi-channel lock-in ampliﬁer coupled to a

*Corresponding author. Address: Department of Applied Physics and Chemistry and Institute for Laser Science, University of Electro-Communications, 1-5-1 Chofugaoka, Chofu, Tokyo 182-8585, Japan. Fax: +81 3 5841 4165.

E-mail addresses:[email protected],[email protected](T. Kobayashi).

Chemical Physics Letters 484 (2010) 354–357

Contents lists available atScienceDirect

Chemical Physics Letters

(2)

polychromator (300 grooves/mm, blazed at 500 nm), which was utilized to disperse the probe pulse and guide it to avalanche pho-todiodes via a 128-channel ﬁber bundle. The spectral resolution of the total system was 1.57 nm. The transmitted probe spectrum after the sample without injection of the pump was recorded by averaging 50 000 laser shots. The transmission difference of the probe was accumulated every 3000 shots under excitation from the pump pulse.

3. Results and discussion

3.1. Pump–probe experimental results

Fig. 3shows the traces of absorbance change,DA(

x

), for deuter-ated indigo carmine; see Ref.[14]for theDA(

x

) spectra of the non-deuterated system. The positiveDA(

x

) in the range longer than 705 nm is dominated by the induced absorption from the lowest excited state. Therefore, the real-time vibrational frequency of the excited molecule can be observed with a minimum contribu-tion of the ground-state wavepacket in this spectral range, if the modulation due to the refractive index induced by the ground-state wavepacket is small.

3.2. Spectrogram analysis

Fig. 4shows a spectrogram[16]obtained by a sliding-window Fourier transform around 720 nm with a Blackman window func-tion of full width at half maximum (FWHM) of 120 fs:

Sð

x

;

s

Þ ¼ Z 1 0 SðtÞgðt

s

Þ expði

x

tÞdt; gðtÞ ¼ 0:42 0:5 cosð2

p

t=TÞ þ 0:08 cosð2

p

t=TÞ; ð1Þ

where T is the gate width of the Blackman window function.

The spectrogram was calculated after applying a high-pass filter to the signal, because slow modulation on the signal (slower than several tens of fs) is caused by fluctuation of laser power instead of molecular vibration. The obtained spectrogram clarifies the change in the instantaneous frequency and the intensity of the vibrational modes during the proton transfer reaction of photo-excited indigo carmine.

Observation of the vibrational coherence after some relaxation or reaction processes may sound hardly understandable. According to the discussions in Refs.[17–20], however, the transfer of vibra-tional coherence and even the creation of coherence by a chemical reaction account for the modulation of transition probability ob-served in an electronic curve crossing process[17]. The modulation of transition probability is thought to be due to the motion of a wavepacket composed of coherent superposition of vibrational eigenstates with different quanta. The bare electronic coupling

[17] and diabatic vibrational frequency have comparable magni-tudes causing considerable interfusion of non-resonant diabatic states. Therefore, the vibrational coherence can be transferred into the product via an electronic curve crossing process under sufﬁ-ciently strong electronic coupling[18–20]. Champion et al. showed experimentally the coherence generated by dissociation.

A peak around 1700 cm1_{, due to the C}_{@O stretching mode}

(

m

C@O) of the reactant[21–23], appeared just after the excitation.

Both of the samples, with and without deuteration, showed that a new peak appeared around 1250 cm1 _{after photo-excitation,}

reﬂecting the formation of the mono-alcohol intermediate

[24,25]. These results agree well with the fact that the Raman fre-quencies of the C–O stretching modes (

m

C–O) of C6H5OH and

C6H5OD have the same value, 1250 cm1 [26]. As for indigo

car-mine, the peak of 1250 cm1_{appeared around 270 fs delay time,}

while it appeared around 410 fs for deuterated indigo carmine. Peaks near 3450 cm1 ₍_{Fig. 4}_{a) and 2380 cm}1 ₍_{Fig. 4}_b)

ap-peared just after photo-excitation and were assigned to the N–H

Fig. 2. (a) Molecular structures of deuterated indigo carmine. (b) Sub-5-fs laser spectrum. (c) Intensity autocorrelation trace (black line) after pulse compression. A ﬁt assuming a Gaussian function yields a pulse width of 7.25 fs (red line). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

Fig. 1. Proton-transfer process in a photo-excited indigo carmine molecule (R = SO3Na).

(3)

stretching (

m

N–H) [22,23] and N–D stretching (

m

N–D) [23]modes,

respectively. The frequency of

m

N–Dis lower by a factor of (2)1/2

than that of non-deuterated indigo carmine. The frequencies of

m

N–Hand

m

N–Dshowed a gradual red shift, and another peak

ap-peared around 3200 and 2250 cm1_{at 270 and 410 fs,}

respec-tively. These peaks located at 3200 and 2250 cm1_{are assignable}

to the O–H (

m

O–H) and O–D (

m

O–D) stretching modes [27–30],

respectively, because these new peaks appeared at nearly the same time as

m

C–O.

3.3. Reaction rates

We analyzed the data to read the probe delay time when the intensities of the

m

C–O,

m

C@O,

m

O–H (

m

O–D) and

m

N–H (

m

N–D) modes

show their peak values. Their delay times were 320, 340, 340 and 330 fs, respectively, and 490, 510, 510, and 520 fs for the corresponding deuterated sample. Therefore, the reaction times in the non-deuterated (deuterated) samples were calculated to be 330 ± 10 (500 ± 13) fs. The latter value gives the reaction rate in deuterated indigo carmine as (2.0 ± 0.1) 1012s1_{in contrast}

with (3.0 ± 0.1) 1012_s1_{in the non-deuterated system. The}

pri-mary KIE obtained from the experimental results predicts that

the proton-transfer rate in the deuterated system is 44% slower than that in the non-deuteride system. The corresponding rate ratio (kH_a=kD_a) between H and D ranges widely from 1.15 (in the early transition state) to several hundred (in the late transi-tion state) [31,32]. Photo-excitation of indigo carmine is ex-pected to proceed through an early transition state due to strong hydrogen bonding, and the observed kHa=k

D

a ratio, 1.5,

ex-ceeds that for the early transition state. This is due to the in-crease in kH_a, indicating that the proton transfer takes place via the tunnel mechanism.

Fig. 3a shows that

m

C@Ois gradually red shifted from 1700 to

1250 cm1_{. On the other hand,}_{Fig. 3}_{b shows that the frequency}

of

m

C@Ois red shifted from 1700 to 1500 cm1, followed by a blue

shift to 1600 cm1_{, and then a red shift to 1250 cm}1_{. This slow}

oscillatory modulation of the

m

C@Ofrequency observed for the

deu-terated system is considered to be induced by the modulation of instantaneous frequency of

m

C@Oby low-frequency modes. These

low-frequency modes are probably coupled to those related to C@C, such as the scissoring modes of C10

@C1

—C2ð

a

1Þ and

C10@C1

—N ð

a

2Þ associated with the deuteron transfer. The

mecha-nism of the mode coupling inducing the modulation can be ex-plained as follows.

Fig. 3. (a) A two-dimensional display of the absorbance change on the probe delay time and the wavelength following the sub-5-fs pulse excitation of deuterated indigo carmine. (b) Dependence of absorbance change on the probe delay time at 720 nm.

Fig. 4. Spectrogram of real-time vibrational modulation of the absorbance change with a Blackman window function of FWHM of 120 fs (a) in indigo carmine shown in Ref. [14]and (b) in deuterated indigo carmine obtained in this work.

(4)

In the ﬁrst deuteron transfer, the distance between N–D and C@O decreases for the deuteron to be transferred to the acceptor. In association with these processes, the C10@C1

—C2 and C10

@C1_{—N bond angles are expected to decrease and increase,}

respectively, with decrease in the carbon–carbon bond order changing from C10@C1

to C10—C1. These changes in the bond angles in turn trigger the scissoring modes of C10

@C1_—C2_{and C}10

@C1_—N.

On the other hand, the oscillatory feature of

m

C@Ohas not been

ob-served in the non-deuterated sample. It is probably caused by a much faster proton-transfer rate than the period of molecular vibrations.

In the deuterated system, the transition state of the recovery process (TSr) of indigo carmine has shorter duration time than that

of the mono-alcohol generation (TSf). This difference may be

as-cribed to that in their energetics as follows: The mono-alcohol sys-tem is unstable and the activation energy to generate indigo carmine (TSr) is lower than that to generate the mono-alcohol

(TSf). Moreover, the reaction pathways of the two processes are

not simply the reversal from forward to backward along the same reaction coordinate space. It is because the reactant of the forward reaction (i.e., deuterated indigo carmine just after photo-excita-tion) is different from the product of the backward reaction (the product after the deuteron transfer). In the non-deuterated system, the intensity of

m

N–Hhas not been observed to decrease around

600 fs. It is because the

m

N–Hmode of 8-fs period is hardly

detect-able by using the sub-5-fs pulse as the intensity of this mode is weakened by decoherence.

4. Conclusion

KIE has been observed directly in the reaction dynamics re-vealed by real-time vibrational spectroscopy using a sub-5-fs pulse, which has enabled us to measure the vibrational response directly and accurately. Direct observation of real-time vibrational spectroscopy using an ultrashort pulse has been proved to be a new powerful method for studies of KIE.

Acknowledgments

This work was supported by a Grant-in-Aid for JSPS Fellows to I.I., a grant to A.Y. from the NSC of Taiwan (NSC

98-2112-M-009-001-MY3), and the grant MOE ATU Program in NCTU to A.Y. and T.K.

References

[1] S.K. Saha, S.J. Maniscalco, H.F. Fisher, Biochemistry 35 (1996) 16483. [2] I.E. Lehoux, B. Mitra, Biochemistry 38 (1999) 5836.

[3] A. Cherrnyshev, R. Pomès, S. Cukierman, Biophys. Chem. 103 (2003) 179. [4] T. Ishimoto, M. Tachikawa, H. Tokiwa, U. Nagashima, Chem. Phys. 314 (2005)

231.

[5] J.J. Gajewski, N.D. Conrad, J. Am. Chem. Soc. 101 (1979) 2747.

[6] L. Kupczyk-Subotkowska, W.H. Saunders Jr., H.J. Shine, W. Subotkowski, J. Am. Chem. Soc. 115 (1993) 5957.

[7] L. Kupczyk-Subotkowska, W.H. Saunders Jr., H.J. Shine, W.J. Subotkowski, Am. Chem. Soc. 116 (1994) 7088.

[8] M.M. Davidson, I.H. Hillier, J. Phys. Chem. 99 (1995) 6748. [9] J.J. Gajewski, Acc. Chem. Res. 30 (1997) 219.

[10] M.P. Meyer, A.J. DelMonte, D.A. Singleton, J. Am. Chem. Soc. 121 (1999) 10865. [11] D.A. Singleton, A.A. Thomas, J. Am. Chem. Soc. 117 (1995) 9357.

[12] B.R. Beno, K.N. Houk, D.A. Singleton, J. Am. Chem. Soc. 118 (1996) 9984. [13] A.J. DelMonte, J. Haller, K.N. Houk, K.B. Sharpless, D.A. Singleton, T. Strassner,

A.A. Thomas, J. Am. Chem. Soc. 119 (1997) 9907.

[14] I. Iwakura, A. Yabushita, T. Kobayashi, Chem. Lett. 38 (2009) 1020. [15] A. Baltuska, T. Fuji, T. Kobayashi, Opt. Lett. 27 (2002) 306.

[16] M.J.J. Vrakking, D.M. Villeneuve, A. Stolow, Phys. Rev. A 54 (1996) R37. [17] J.M. Jean, G.R. Fleming, J. Chem. Phys. 103 (1995) 2092.

[18] F. Rosca, A.T.N. Kumar, X. Ye, T. Sjodin, A.A. Demidov, P.M. Champion, J. Phys. Chem. A 104 (2000) 4280.

[19] A.T.N. Kumar, F. Rosca, A. Widom, P.M. Champion, J. Chem. Phys. 114 (2001) 701.

[20] P.M. Champion, F. Rosca, D. Ionascu, W. Cao, X. Ye, Faraday Discuss. 127 (2004) 123.

[21] I.T. Shadi, B.Z. Chowdhry, M.J. Snowden, R. Withnall, Chem. Commun. (2004) 1436.

[22] R. Giustetto, F.X. Llabrés i Xamena, G. Ricchiardi, S. Bordiga, A. Damin, R. Gobetto, M.R. Chierotti, J. Phys. Chem. B (2005) 19360.

[23] T.M. Ivanova1, G.M. Ostapchuk1, L.B. Shagalov1, V.N. Berdyugin, Khim. Geterotsikl. Soedin. (1981) 501.

[24] W. Lüttke, M. Klessinger, Chem. Ber. 97 (1964) 2342. [25] R.L. Hinman, J. Org. Chem 29 (1964) 1449.

[26] G. Keresztury, F. Billes, M. Kubinyi, T. Sundius, J. Phys. Chem. A 102 (1998) 1371.

[27] M.A. Habib, J.O.M. Bockris, Langmuir 2 (1986) 388.

[28] H. Kandori, M. Belenky, J. Herzfeld, Biochemistry 41 (2002) 6026. [29] K. Shimono, Y. Furutani, N. Kamo, H. Kandori, Biochemistry 42 (2003) 7801. [30] M. Sovago, R.K. Campen, H.J. Bakker, M. Bonn, Chem. Phys. Lett. 470 (2009) 7. [31] A. Chernyshev, R. Pomès, S. Cukieran, Bio. Chem. 103 (2003) 179.

[32] U. Maitra, J. Chandrasekhar, Resonance 2 (1997) 18.