Thermolysis and photolysis of 2-ethyl-4-nitro-1(2H)-isoquinolinium hydroperoxide

Research Article

Received: 20 December 2012,

Revised: 5 February 2013,

Accepted: 8 February 2013,

Published online in Wiley Online Library: 1 April 2013

(wileyonlinelibrary.com) DOI: 10.1002/poc.3107

Thermolysis and photolysis of 2-ethyl-4-nitro-1

(2H)-isoquinolinium hydroperoxide

Renat Khatmullin, Dapeng Zhou, Thomas Corrigan,

Ekaterina Mirzakulova and Ksenija D. Glusac*

The thermal and light-induced O ꢀ O bond breaking of 2-ethyl-4-nitro-1(2H)-isoquinolinium hydroperoxide (IQOOH)

were studied using ¹H NMR, steady-state UV/vis spectroscopy, femtosecond UV/vis transient absorption (fs TA) and

time-dependent density functional theory (TD DFT) calculations. Thermal O ꢀ O bond breaking occurs at room

temperature to generate water and the corresponding amide. The rate of this reaction, k = 5.4 ꢁ 10^ꢀ6

s

^ꢀ1, is higher

than the analogous rates of simple alkyl and aryl hydroperoxides; however, the rate signiﬁcantly decreases in the

presence of small amounts of methanol. The calculated structure of the transition state suggests that the thermolysis

is facilitated by a 1,2 proton shift. The photochemical process yields the same products, as conﬁrmed using NMR and

UV/vis spectroscopy. However, the quantum yield for the photolysis is low (Φ = 0.7%). Fs TA studies provide

additional detail of the photochemical process and suggest that the S₁state of IQOOH undergoes fast internal

conversion to the ground state, and this process competes with the excited-state O ꢀ O bond breaking. This result

was supported by the fact that the model compound IQOH exhibits similar excited-state decay lifetimes as IQOOH,

Supporting information may be found in the online version of this paper.

Keywords: organic hydroperoxide; thermolysis; photolysis; pump probe spectroscopy

The mechanism of hydroperoxide thermolysis is sensitive to the

INTRODUCTION

initial hydroperoxide concentration. At low concentrations,

decomposition follows the ﬁrst-order kinetics, while at higher

concentrations the switch to a bimolecular mechanism occurs.

Hydrogen-bonded hydroperoxide dimers are proposed to play an

important role in the bimolecular mechanism.^[32,33]The solvent

characteristics play a signiﬁcant role in the decomposition of

organic hydroperoxides. Speciﬁcally, the purity of the solvent^[19,25]

is vital, since traces of strong acids and metallic complexes can cat-

alyze the O –O bond breakage. Furthermore, the rate of thermolysis

was found to be signiﬁcantly higher in alcoholic and unsaturated

solvents than in aromatic or chlorinated hydrocarbon solvents.^[28]

These results are explained in terms of a radical forming, bimolecu-

lar solvent (SH)/hydroperoxide (ROOH) thermolysis of the type:

ROOH + SH ! RO^.+ S^.+ H₂O. The presence of radical-forming

substances can also lead to the suppression of the hydroperoxide

decomposition, if the formed radicals S^.are chemically inert. These

“antioxidants” are usually phenol and aniline derivatives and lead to

a decrease of the hydroperoxide decomposition rates.^[29–31,34]

Hydroperoxides that exhibit a heteroatom in the a-position

display a special type of reactivity, which is proposed to occur

by the heterolytic O – O bond breaking in the presence of

Most organic peroxides are thermally and photochemically unsta-

ble, due to the presence of a weak oxygen–oxygen bond.^[1,2]The

cleavage of the O –O bond in peroxides usually requires ~34 kcal/

mol of input energy, and this low bond strength is caused by the

nonpolar character of the O – O bond (energies of polar bonds,

such as those made of heteroatoms with sufﬁciently different elec-

tronegativities are larger than the energies of nonpolar bonds; for

example, the bond energy of N –O bond is 53 kcal/mol, while the

N–N bond energy is 38 kcal/mol)^[3]and destabilization of the O –O

bond due to the repulsion by nonbonding electron pairs of the two

oxygen atoms. Thus, the photochemical and thermal decomposition

of organic peroxides is generally initiated by the cleavage of O –O

bond to form oxygen-centered radicals, which undergo additional

chemical reactions, such as hydrogen atom abstraction,^[4–6]

b-scission,^[6]decarboxylation,^[7–9]rearrangements,^[10,11]and addition

to multiple carbon–carbon bonds.^[12,13]

Organic hydroperoxides are formed and decomposed during

the oxidation of organic compounds by molecular oxygen, which

is the leading source of concern for the air stability of food,

gasoline, plastics and other organic materials.^[14–18]For this reason,

decomposition of organic hydroperoxides has been addressed in a

large number of reports,^[19–26]all of which have clearly indicated

the high complexity of the reaction mechanism. The decomposi-

tion usually begins with the O – O bond homolysis of the hydroper-

oxide, followed by a chain of reactions involving hydroxyl and

alkoxyl radicals.^[18]The reaction mechanism, rates and product

distribution are readily affected by experimental conditions, such

as the concentration of the hydroperoxide,^[25]temperature,^[25,27]

solvent^[28]and radical initiators^[21,28]or inhibitors.^[29–31]

* Correspondence to: Ksenija D. Glusac, Department of Chemistry, Center for

Photochemical Sciences, Bowling Green State University, Bowling Green, OH

43403, USA.

E-mail: kglusac@bgsu.edu

a R. Khatmullin, D. Zhou, T. Corrigan, E. Mirzakulova, K. D. Glusac

Department of Chemistry, Bowling Green State University, Bowling Green, OH,

43403, USA

J. Phys. Org. Chem. 2013, 26 440–450

THERMOLYSIS AND PHOTOLYSIS OF ISOQUINOLINIUM HYDROPEROXIDE

nucleophiles (S_N2 mechanism).^[35]This polarization of the O – O

bond enables the a-substituted hydroperoxides to incorporate

atomic oxygen into organic substrates, such as amines, sulﬁdes, al-

kenes, alcohols and ketones.^[36–42]The a-heteroatom is in most

cases oxygen (as in a-hydroperoxyethers^[40]or peracids^[43,44]) or ni-

trogen (as in a-aminohydroperoxides^[40]or a-azo hydroperox-

ides^[36,45,46]), and, in some cases, the insertion of oxygen is believed

to involve an intramolecular H-bond between the hydroperoxylic

proton and the a-heteroatom.^{[36,43,45,46]}The oxygen insertion can

also be made catalytic if the alpha-substituted hydroperoxide is

obtained from iminium ions that reversibly form pseudobase

intermediates.^{[38,42,47–52]}

This manuscript reports the room-temperature (RT) thermolysis

of one such heterocyclic hydroperoxide, derived from the

isoquinolinium ion derivative. In speciﬁc, as part of our ongoing

interest in organocatalytic water oxidation by iminium ions,^[53,54]

the isoquinolinium hydroperoxide (IQOOH) was prepared (Fig. 1),

and its thermal and photochemical behavior was studied using

steady-state UV/vis and NMR spectroscopy, femtosecond transient

absorption (fs TA) laser spectroscopy and density functional theory

(DFT) calculations. It was found that IQOOH is, in aprotic solvents,

less stable than other known hydroperoxides that lack the alpha

heteroatom and that the O – O bond breaking in IQOOH occurs

thermally at RT to yield 2-ethyl-4-nitro-1(2H)-isoquinolinone (IQ

amide). Interestingly, the stability of IQOOH is signiﬁcantly

increased in the presence of small amounts of methanol, and the

possible reasons for this stabilization effect are discussed. The

photoinduced O – O bond breaking of IQOOH generates the same

product, but with a surprisingly low quantum yield. These results

are discussed in terms of a fast competing internal conversion to

the IQOOH ground state via an S₁–S₀conical intersection.

with concentrated aqueous ammonia, the resulting solution

was extracted with chloroform (3 ꢃ 150 ml). Combined extracts

were washed with 1M HCl (4 ꢃ 100 ml) and dried over anhydrous

calcium chloride. After removal of the solvent under reduced

pressure, the residue was dissolved in ether. Hydrogen-chloride

gas was bubbled through the ethereal solution to precipitate

4-nitroisoqunolinium chloride. The resulting precipitate was

ﬁltered and dissolved in distilled water. An aqueous solution of this

salt was extracted with chloroform. The organic solution was dried

over calcium chloride, and the solvent was evaporated in vacuo to

give 0.73 g (5.4%) of 4-nitroisoquinoline. MS ESI⁺, m/z: found, 174;

calculated for C₉H₆O₂N₂⁺, 174.16. ¹H NMR (300 MHz, CDCl₃) d ppm:

9.46 (d, J = 0.8 Hz, 1H), 9.31 (s, 1H), 8.69 (dd, J = 8.7, 1.0 Hz, 1H), 8.15

(dt, J = 8.2, 1.0 Hz, 1H), 7.99 (m, J = 8.5, 7.0, 1.4 Hz, 1H), 7.80 (m,

J = 8.1, 7.0, 1.0 Hz, 1H); ¹³C NMR (75 MHz, CDCl₃) d ppm: 158.0,

141.0, 137.4, 134.2, 129, 128.8, 128.3, 127.8, 122.5.

2-ethyl-4-nitroisoquinolinium tetraﬂuoroborate (IQ⁺)

IQ⁺was prepared by a procedure similar to that described by

Lorance et al.^[56]Triethyloxonium tetraﬂuoroborate (400 mg, 2.1

mmol) was added to a 20 ml solution of 4-nitroisoquinoline

(290 mg, 1.7 mmol) in dichloromethane. Mixture was allowed

to stir for 4 h at RT. The resulting precipitate was ﬁltered and

recrystallized from methanol to give 0.23 g (51%). MS ESI⁺, m/z:

found, 203; calculated for C₁₁H₁₁O₂N⁺₂, 203.22. ¹H NMR (300

MHz, CD₃CN-d3), d ppm: 9.87 (s, 1H), 9.35 (d, J = 1.3 Hz, 1H),

8.81 (d, J = 8.8 Hz, 1H), 8.61 (dd, J = 8.2, 1.2 Hz, 1H), 8.48

(m, J = 8.5, 7.0, 1.3 Hz, 1H), 8.27 – 8.15 (m, 1H), 4.81 (q, J = 7.3

Hz, 2H), 1.71 (t, J = 7.3 Hz, 3H); ¹³C NMR (75 MHz, CD₃CN-d3)

d ppm: 153.88, 140.68, 133.61, 132.70, 131.90, 130.45, 129.10,

123.19, 117.78, 57.84, 15.19.

EXPERIMENTAL PART

2-ethyl-4-nitro-1(2H)-isoquinolinium hydroperoxide (IQOOH)

Finely crushed ureaꢁhydrogen peroxide tablets (7 g, 35% hydrogen

peroxide, 72.06 mmol H₂O₂) were added to a 15 ml solution of IQ⁺

(0.6 g, 2.07 mmol) in methanol. The resulting mixture was wrapped

with aluminum foil and stirred for 20 min at RT (298K) in the

absence of light. Methanol was evaporated, and the remainder

was puriﬁed by column chromatography using ethyl acetate/cy-

clohexane mixture (3:2) as an eluent to give 0.14 g (30%) of IQOOH.

MS MALDI⁺, m/z: found, 236; calculated for C₁₁H₁₂O₄N⁺₂, 236.22. ¹H

NMR (300 MHz, CD₃CN-d3), d ppm: 9.51 (d, J = 0.6 Hz, 1H), 8.67 (d,

J = 1.2 Hz, 1H), 8.61 – 8.42 (m, 1H), 7.84 – 6.98 (m, 3H), 6.34 (d, J = 1.3

Hz, 1H), 4.30 – 3.37 (m, 2H), 1.40 (t, J = 7.3 Hz, 3H); ¹³C NMR (75 MHz,

CD₃CN-d3) d ppm: 146.78, 130.88, 129.70, 128.10, 127.67, 127.51,

123.99, 122.73, 91.22, 50.60, 14.86.

General methods

All starting compounds and solvents were purchased from

Sigma-Aldrich and used as received. The thermal decomposition

of IQOOH was studied at RT (298 K). Solvents used prior to kinetic

measurements were carefully distilled several times over CaH₂.

¹H and ¹³C NMR spectra were recorded on Bruker Avance 300

MHz and Bruker Avance 3 500 MHz system. MS-EI spectra were

measured on Shimadzu GC-MS-Q5050A spectrometer.

SYNTHESIS

4-nitroisoquinoline

Synthesis of 4-nitroisoquinoline was described previously.^[55]

Concentrated nitric acid (5 ml, 84 mmol, 70%) was slowly added

to a solution of 10.0 g (77 mmol) of isoquinoline in 100 ml of

acetic anhydride. The solution was heated at 80 ^ꢂC for 12 h

and then poured over 300 g of crushed ice. After neutralization

2-ethyl-4-nitro-1(2H)-isoquinolinone (IQ amide)

IQ amide was obtained by thermal decomposition of IQOOH (0.2

g, 0.86 mmol) in 25 ml of acetonitrile at RT for 10 days. The

solvent was evaporated, and the residue was recrystallized from

Figure 1. Synthetic procedure for preparation of IQ⁺and IQOOH

J. Phys. Org. Chem. 2013, 26 440–450

wileyonlinelibrary.com/journal/poc

R. KHATMULLIN ET AL.

diethyl ether to give 0.08g (43%). MS ESI⁺, m/z: found, 219; calcu-

DFT calculations

1

lated for C₁₁H₁₁O₃N⁺₂, 218.21. H NMR (300 MHz, CD₃CN-d3), d

All calculations were performed at the Ohio Supercomputer

Center. Ground and transition state (TS) optimization of IQOOH

in the gas phase were performed using Gaussian 09 software

at the B3LYP/6-31 + G** level^[61,62]The vibrational frequency

calculations (B3LYP/6-31 + G**) were used to conﬁrm that all

stationary points were energy minima (no imaginary vibrational

frequencies). Difference density (DD) plots were generated from

the computed electron densities of the ground and excited states

(TD-B3LYP/6-31 + G**) using Gaussian 09 with consideration of

implicit solvation of acetonitrile (using the polarization continuum

model, PCM).^[63–68]

ppm: 8.81 (s, 1H), 8.64 (m, J = 8.5, 1.1, 0.6 Hz, 1H), 8.42 (m,

J = 8.1, 1.5, 0.6 Hz, 1H), 8.11 – 7.74 (m, 1H), 7.66 (m, J = 8.2, 7.2,

1.1 Hz, 1H), 4.13 (q, J = 7.2 Hz, 2H), 1.38 (t, J = 7.2 Hz, 3H).

Steady-state measurements

Absorption spectra were recorded on Varian Cary 100 BIO UV spec-

trometer in 1 cm pathlenght quartz cell. Fluorescence spectra were

recorded on Edinburg Instruments Fluorimeter equipped with

Xe900 lamp in 1 cm pathlenght quartz cell. Tris(2,2’-bipyridine)

ruthenium bis(hexaﬂuorophosphate) in acetonitrile (Φ = 0.059)

was used as a standard for ﬂuorescence quantum yield.^[57]The

solutions for ﬂuorescence measurements were prepared with

absorption of 0.1 at the excitation wavelength. FTIR spectra were

collected using FTIR-8400S (Shimadzu).

RESULTS AND DISCUSSION

Model compound

This study investigates the behavior of a model hydroperoxide

IQOOH derived from an isoquinolinium ion IQ⁺(Fig. 1). The

IQ⁺/IQOOH system exhibits the key characteristics similar to

ﬂavinium and other iminium ions found to catalytically insert

oxygen atom into organic substrates,^[47–51]such as: (i) IQ⁺reacts

with H₂O₂to generate IQOOH, which can potentially oxidize sul-

ﬁdes, amines and other organic substrates by a heterolytic O – O

bond breakage (i.e. IQOOH + R₂S ! IQOH + R₂S = O); (ii) Since the

IQOH pseudobase pKa value is 5.03,^[69]IQ⁺can easily be

regenerated from IQOH by the addition of a weak acid.

Even though this study does not address directly the catalytic

oxidation reactions by IQOOH, it provides useful information

regarding the O – O bond breaking pathways in these types of

compounds. The hydroperoxide, IQOOH, was synthesized from

isoquinoline 1 according to Fig. 1. The nitration of isoquinoline

1 to generate 4-nitroisoquinoline 2 was achieved using nitric

acid and acetic anhydride according to the method described

by Bunting et al.^[55]Yield of 2 was low, at merely 6%, due to

the lack of reactivity of the pyridine moiety of isoquinoline

towards electrophilic aromatic substitution. Despite the low yield

of this reaction, the nitration of isoquinoline was performed in

order to increase the electrophilicity of the C1 atom in the ring

and thus promote the formation of the hydroperoxide IQOOH.

The N-alkylation of compound 2 using triethyl oxonium

tetraﬂuoroborate generated N-ethyl-4-nitroisoqunolinium salt

(IQ⁺), which was further treated with H₂O₂to produce IQOOH.

RT thermolysis, photolysis and quantum yield experiments

The thermal bond breakage of IQOOH was performed using 1 ml

of a 0.02 M solution of IQOOH in acetonitrile. The solution was

kept at RT (298 K) in the absence of light until full decomposition

was achieved. Light-induced bond breaking was carried out in a

4 ml quartz cuvette with a 0.02 M solution of IQOOH in acetoni-

trile. Radiation source was a 400 W Hg vapor quartz immersion

lamp. A 400 nm short-pass colored glass ﬁlter was used to

prevent irradiation above the desired wavelength. Dry ice was

used to avoid thermal decomposition due to heat emitted from

the lamp. Potassium ferrioxalate was used as a standard (Φ_S=

1.14 at 405 nm) to measure the quantum yield of IQOOH

photolysis (Φ_IQOOH).^[57,58]

TA measurements

The laser system for the ultrafast TA measurement was described

previously.^[59]Brieﬂy, the 800 nm laser pulses were produced at a

1 kHz repetition rate (fwhm = 110 fs) by a mode-locked Ti:sap-

phire laser (Hurricane, Spectra-Physics). The output from a Hurri-

cane was split into pump (85%) and probe (10%) beams. The

pump beam (800 nm) was sent into a second harmonic genera-

tor (Super Tripler, CSK) to obtain a 400 nm excitation source. The

energy of the pump beam was 2 mJ/pulse. The probe beam was

focused into a horizontally translated CaF₂crystal for white light

continuum generation between 350 and 800 nm. The ﬂow cell

(Starna Cell Inc. 45-Q-2, 0.9 ml volume with 2 mm pathlenght),

pumped by a Fluid Metering RHSY Lab pump (Scientiﬁc Support

Inc.), was used to prevent photodegradation of the sample. After

passing through the cell, the continuum was coupled into an

optical ﬁber and input into a CCD spectrograph (Ocean Optics,

S2000). The data acquisition was achieved using in-house

LabVIEW (National Instruments) software routines. The group

velocity dispersion of the probe pulse was determined using

nonresonant optical Kerr effect (OKE) measurements.^[60]Sample

solutions were prepared at a concentration needed to have

absorbance of A ~ 0.8–1 at the excitation wavelength. Data pro-

cessing and analysis was achieved as follows: (i) Chirp correction

was determined using nonresonant OKE measurements. (ii)

Noise was reduced using a singular value decomposition

method (Matlab 7.1). (iii) Data were analyzed using SPECFIT/

32^TMGlobal Analysis System (Spectrum Software Associates,

MA, USA).

RT thermolysis of IQOOH

The solution of IQOOH in deuterated acetonitrile was found to

undergo thermal decomposition at RT (298K) within 4 days.

The reaction was monitored using ¹H NMR spectroscopy (Fig. 2).

The NMR spectrum of IQOOH (Fig. 2a) exhibits two characteristic

peaks: a multiplet at 3.8 ppm corresponding to the two

diastereotopic methylene protons of the N-ethyl group, and a

singlet at 9.5 ppm corresponding to the proton of the

hydroperoxy group. These peaks are diminished as a function

of time (Fig. 2a and 2b), and the appearance of new signals at

7.8 and 8.4–8.8 ppm is observed. After RT thermolysis was com-

1

pleted, the product was isolated and characterized by H NMR

spectroscopy (Fig. 2c) as IQ amide (Fig. 3). IQ amide has speciﬁc

NMR signals with chemical shifts at 4.15 ppm (quartet) due to

H-atoms on carbon 1 and a singlet at 8.8 ppm due to the H-atom

on carbon 6. Additional NMR peaks were observed, which are

formed and diminished during the course of the reaction. These

wileyonlinelibrary.com/journal/poc

J. Phys. Org. Chem. 2013, 26 440–450

THERMOLYSIS AND PHOTOLYSIS OF ISOQUINOLINIUM HYDROPEROXIDE

peaks do not correspond to the starting hydroperoxide or the IQ

amide product and could belong to either the reaction intermediate

or the unstable byproduct.

One possible mechanism for IQ amide formation involves the

homolytic O – O bond breaking of IQOOH, followed by a hydrogen

atom abstraction via hydroxyl radical: IQOOH! IQO^.+ OH ! IQ

.

Figure 3. Formation of IQ amide upon RT thermolysis of IQOOH

amide + H₂O. This mechanism is supported by the fact that the

homolytic O – O bond scission is known to occur in many hydro-

peroxides.^{[25,30,70,28]}However, closer inspection of the reported

rates of hydroperoxide thermolysis show that the O– O bond

breaking usually occurs at a much slower rate than observed in this

study. For example, Table 1 lists the ﬁrst-order homolysis rates of

several alkyl and aryl hydroperoxides calculated using the previ-

ously reported activation energies and pre-exponential

slow down the RT thermolysis. This behavior was indeed

observed, as shown in Fig. 4.

The progress of IQOOH RT thermolysis was studied in the

absence (Fig. 4a) and presence (Fig. 4b) of methanol using UV/

vis absorption spectroscopy, by monitoring the disappearance

of the IQOOH absorption band at l_max= 384 nm and the growth

of the IQ amide absorption at 360 nm. In the absence of metha-

nol, signiﬁcant amount of IQOOH converted into IQ amide within

28 h. In the presence of only 2% of methanol, the stability of

IQOOH is signiﬁcantly increased, and essentially no RT

thermolysis is observed over the period of 5 days. Even though

this preliminary study on methanol effect seems to support the

presence of hydrogen-bonded dimers, the follow-up experi-

ments have ruled this mechanism out. For example, the dimer

mechanism should exhibit the second-order kinetics with

respect to IQOOH concentration. To the contrary, the kinetic

studies on IQOOH in acetonitrile (Fig. 5) suggest that the RT

thermolysis is ﬁrst order with respect to IQOOH concentration,

factors.^[25,30]At RT, the calculated rates fall within the 10^ꢀ16

–

10^ꢀ10s^ꢀ1ranges, suggesting that no observable decomposition

of these compounds is observed at RT within the 4 days period.

It is important to note that we performed careful puriﬁcation of

the solvent, to ensure that the traces of impurities, such as metal

ions or acids, do not catalyze the observed reaction.

One possible cause for high IQOOH decomposition rate is that

the presence of the a-nitrogen atom in IQOOH decreases its O –O

bond dissociation energy (BDE), presumably by providing the

extended delocalization of the spin density of IQO^.radicals. We used

DFT at the B3LYP/6-31 + G** level to evaluate BDEs of IQOOH and

several other hydroperoxides (Table 2). The calculated BDEs for

MeOOH and tBuOOH are lower than the experimentally obtained

values,^[71,72]and this effect was also observed previously by Bach

et al.^[73]The BDE for IQOOH was found to be 35.7 kcal/mol.

While this value is by 1 kcal/mol lower than the BDEs of alkyl

hydroperoxides, it is not sufﬁcient to explain the differences in the

decomposition rates.

with the rate constant of k ~ 5.4 ꢁ 10^ꢀ6

s

^ꢀ1. This value is signiﬁ-

cantly lower than the estimated RT rate constant for several

other hydroperoxides (Table 1).

Previous reports have shown that, in some cases, the

thermolysis of hydroperoxides occurs faster than expected

based on their O – O bond dissociation energies.^[32,33]In these

studies, the hydrogen-bonded hydroperoxide dimers were pos-

tulated to be involved in the homolysis step. To investigate the

role of hydrogen-bonded dimer formation in the case of IQOOH,

we investigated the effect of methanol on the rate of IQ amide

formation. Due to its tendency to form hydrogen bonds, metha-

nol is expected to break the hydroperoxide dimer and possibly

Table 1. Room temperature decomposition rates, frequency

factors and activation energies of several hydroperoxides

ROOH

A, s^ꢀ1

E_akcal/mol

k_estat 25^ꢂCs^ꢀ1

t-Bu^a

1.0 ꢁ 10¹⁶

1.3 ꢁ 10¹¹

1.0 ꢁ 10¹⁰

8.5 ꢁ 10¹³

1.3 ꢁ 10¹¹

43^a

29^b

1.0 ꢁ 10^ꢀ16

1.0 ꢁ 10^ꢀ12

1.7 ꢁ 10^ꢀ10

1.0 ꢁ 10^ꢀ11

5.0 ꢁ 10^ꢀ13

a-cumyl^b

n-octyl^b

decalyl^b

tetralin^b

26.9^b

32.1^b

29^b

^aValues obtained from reference^[25]

^bValues obtained from reference.^[30]

Table 2. O–O Bond dissociation energies ΔH₂₉₈, Gibbs free

energies ΔG₂₉₈and standard entropies of formation ΔS₂₉₈(kcal/

mol) of several hydroperoxides, computed at B3LYP/6-31 + G**

level of theory

ROOH

ΔH₂₉₈(experiment) ΔH₂₉₈Δ G₂₉₈Δ S₂₉₈

MeOOH

t-BuOOH

IQOOH

44.6^a

36.6

36.7

35.7

26.7

26.3

25.1

0.03

43.8^b

-----

^aValues obtained from reference^[71]

^bValues obtained from reference^[72]. All values reported in

kcal/mol. The frequencies were scaled by factor 0.9804

(thermal correction).

Figure 2. Changes in ¹H NMR spectra of 0.015 M IQOOH in CD₃CN after:

a) 0 min; b) 34 h and c) 4 days of RT thermolysis

J. Phys. Org. Chem. 2013, 26 440–450

wileyonlinelibrary.com/journal/poc

R. KHATMULLIN ET AL.

1

The lack of IQOOH dimers was also conﬁrmed using H NMR

reaction, we observed the formation of pseudobase IQOH and the

oxidized dimethyl sulfoxide. These results show that IQOOH also

undergoes the oxygen-atom transfer reactions characteristic of

other a-heteroatom hydroperoxides.

spectroscopy. Figure 6a represents the changes in NMR spectra

of IQOOH upon addition of methanol. For comparison, Fig. 6b

outlines the effect of methanol on IQOH, a model compound

that generates hydrogen-bonded dimers. In case of IQOOH, the

increase of methanol concentration only leads to a monotonous

shift of OOH proton (proton 9) from 9.51 to 10.4 ppm as well as

the broadening of the signal (Fig. 6a). The shift of the OOH

proton to a higher frequency can be explained by intermolecular

hydrogen bonding with methanol molecules. On the other hand,

the effect of methanol on IQOH is much more prominent. Upon

addition of methanol to IQOH, we observe the change in the

chemical shifts of all protons and their coupling constants

(Fig. 6b) that is indicative of the disassembly of IQOH dimers in

the presence of methanol. If IQOOH were present in a similar

dimer network, we would also expect the change in the coupling

constants of all protons, similar to IQOH. Furthermore, the

concentration-dependent NMR studies of IQOOH and IQOH

(Fig. S1, Supporting Information) provide additional evidence of

the absence of IQOOH hydrogen-bonded dimers and the

presence of dimers in the case of IQOH. Based on these NMR

and kinetic ﬁndings, we ruled out the possibility of hydrogen-

bonded IQOOH dimers formation in our studies.

In contrast to the low thermolysis rates observed with alkyl

and aryl hydroperoxides, the hydroperoxides that contain

a-heteroatoms exhibit signiﬁcantly higher rates of thermolysis

(for example, a ﬂavin-based hydroperoxide exhibits the rate

constant at RT of 1.3 ꢁ 10^ꢀ5s^ꢀ1).^[74]In the presence of appropriate

amines or sulﬁdes, the oxygen-atom transfer from hydroperoxide

to the substrate was found to occur,^[75–77]and the mechanism of

this reaction is not well understood. It is proposed to involve a

heterolytic O – O bond dissociation by an S_N2-type reaction of

the type: R₂S + R’OOH ! R₂SO+ R’OH. The role of the heteroatom

in this chemistry is to polarize the O – O bond towards heterolysis.

The support of the S_N2 mechanism is obtained in a study showing

that the rate of oxygen transfer depends on the pKa value of the

formed pseudobase R’OH.^[78]Given the structural similarities

between IQOOH and the ﬂavin hydroperoxide and the fact that

the IQOH pKa value is 11.15^[69](similar to ﬂavin pseudobase,

pKa = 9.9^[79]), we anticipated that IQOOH can perform similar

oxygen atom transfer processes in the presence of suitable

substrates. To conﬁrm this behavior, we investigated a reaction be-

tween dimethyl sulﬁde and IQOOH (Fig. S2, Supporting Material).

In addition to the formation of IQ amide due to the intramolecular

While an S_N2-type reaction can be proposed for a reaction

between dimethyl sulﬁde and IQOOH, this type of chemistry

cannot explain the observed intramolecular RT thermolysis of

IQOOH to generate IQ amide and water. In this case, the

nucleophile (hydride anion) would need to attack the oxygen

atom via strained four-membered TS, which seems energetically

demanding. Furthermore, the S_N2-type mechanism cannot

explain the strong methanol effect observed for IQOOH RT

thermolysis (Fig. 4b). To obtain a better mechanistic understand-

ing of this low-barrier thermolysis, the TS was optimized at the

UB3LYP/6-31 + G** level (Figure 7b). Surprisingly, we located

the TS at only 1.11 kcal/mol above the IQOOH ground state,

and the structure shows clearly that the C1 proton has shifted

to the C9 position. This proton shift is likely facilitated by the

electron donation from the nitrogen lone pair, as shown in

Figure 7a. In addition to the proton shift, the O–O bond is

signiﬁcantly elongated (the TS bond length is 2.478 Å), while

the weak C1–O bond is formed (TS bond length is 1.745 Å). Thus,

a hemiacetal-type TS results, which further undergoes water

elimination to generate the IQ amide, as illustrated in Figure 7a.

Based on this calculation, we propose that the IQOOH

Figure 5. A ﬁrst-order proﬁle of the thermal decomposition of IQOOH

solutions in acetonitrile at room temperature (298K): red (0.1 mM); blue

(0.18 mM)

b

a

8

6

4

2

0

6

4

2

0

300

400

300

400

Wavelength/nm

Figure 4. The changes in UV/vis absorption spectra of (a) 0.15 mM IQOOH in acetonitrile after 5 min (red), 2 (orange), 21 (green) and 28 h (maroon)

and (b) 0.16 mM IQOOH in acetonitrile/methanol mixture (98:2) after 5 min (red), 1 (red), 2 (orange), 3 (green) and 5 days (maroon) at room temperature

thermal decomposition. The blue dashed line represents the UV/vis spectrum of 0.1 mM IQ amide in acetonitrile

wileyonlinelibrary.com/journal/poc

J. Phys. Org. Chem. 2013, 26 440–450

THERMOLYSIS AND PHOTOLYSIS OF ISOQUINOLINIUM HYDROPEROXIDE

a

b

Figure 6. Changes in ¹H NMR spectra upon addition of methanol to: (a) 0.15 mM of IQOOH in CD₃CN; (b) 0.18 mM of IQOH in CD₃CN; (d)-dimer, (m)-monomer

thermolysis is facilitated by the presence of the adjacent

nitrogen center, which, through conjugation with the aromatic

moiety, serves as an internal base to abstract the C–H proton

and facilitate the nucleophilic attack of C1 carbon at the distal

oxygen center of the OOH moiety. This behavior would explain

the discrepancy between the rates reported in Table 1 for several

alkyl hydroperoxides and the rate of IQOOH RT thermolysis. It is

possible that such a mechanism operates in catalytic oxidation

reactions of nucleophiles by other hydroperoxides that exhibit

a heteroatom in the a-position.

activation energy of IQOOH with the barrier found in benzyl

hydroperoxide. Benzyl hydroperoxide is structurally similar to

IQOOH, but lacks the basic nitrogen site, and it requires 41.1

kcal/mol to undergo O–O bond scission and water elimina-

tion.^[80]This comparison shows that the presence of N-center

in IQOOH drastically stabilizes the TS for thermolysis. The 1.11

kcal/mol barrier is very small, suggesting that the reaction likely

occurs in multiple steps, with other TSs of higher energy. Addi-

tional computational studies are underway to fully characterize

the reactivity of IQOOH.

In light of this mechanism, the effect of methanol becomes

clear: hydrogen bonding between methanol and the nitrogen

center of IQOOH is expected to weaken the electron-donating

ability of the nitrogen center, thus raising the barrier for the

C1-H proton abstraction required to reach the TS in Figure 7.

For example, it is interesting to compare the 1.11 kcal/mol

Photolysis of IQOOH

Since RT thermolysis of IQOOH occurs at RT, we postulated that

the O – O bond cleavage might occur with the investment of

light energy as well. The UV/vis absorption and ﬂuorescence

J. Phys. Org. Chem. 2013, 26 440–450

wileyonlinelibrary.com/journal/poc

R. KHATMULLIN ET AL.

Figure 7. (a) Proposed mechanism for the formation of IQ amide upon RT thermolysis of IQOOH. (b) Calculated transition structure (two projections)

for the 1,2-proton transfer of IQOOH at the UB3LYP/6-311 + G^**level

spectra of IQOOH (Fig. 8a) offer some support for the presence of

photolysis: (i) the lowest energy absorption of IQOOH appears at

384 nm (3.23 eV), which is more than sufﬁcient to break the O–O

bond (calculated BDE is 1.55 eV, Table 1); (ii) the IQOOH ﬂuores-

cence (l_max= 514 nm) exhibits very low quantum yield

(Φ = 5.4ꢁ10^ꢀ4), which is plausibly caused by the excited-state

O – O bond cleavage of IQOOH. To perform the initial character-

ization of IQOOH excited states, time-dependent DFT (TD-DFT,

B3LYP/6-31 + G**, PCM acetonitrile) was used to evaluate the ﬁrst

four singlet excited states of IQOOH. The DFT methodology

reproduces the experimental ﬁndings to a satisfactory degree:

(i) calculated ground-state vibrational frequencies match well

with the experimental values, as shown in Fig. S3, Supporting

Information; (ii) calculated electronic transition energies

correlate well with the experimental UV/vis absorption spectrum,

as shown in Figure 8a. Given this good correlation with the

experiment, the results of TD-DFT calculations were used to

generate DD plots presented in Fig. 8b.

The DD plots show sections of the IQOOH molecular framework

in which the electronic charge is depleted (red color) or accumu-

lated (green color) upon excitation. One obvious observation that

can be made from DD plots is that most of the charge redistribu-

tion involves the isoquinoline framework, while the O– O moiety

exhibits only minor changes in the electron density redistribution

upon excitation. The photolysis of peroxides usually occurs via

n ! s^*transition,^[81]and this type of excited state should

give rise to DD plots with large excited-state redistribution of

electronic charge around the O – O group. On the contrary, the

S₁! S₀DD plot of IQOOH exhibits essentially p ! p^*character

with a certain degree of charge-transfer from the donating

phenyl group to accepting nitro group. These results suggest

that the Frank–Condon point of the S₁state of IQOOH does

b)

a)

0.30

1.0

0.8

0.6

0.4

0.2

0.0

0.25

0.20

0.15

0.10

0.05

0.00

S₁

S₄

S₂

S₃

300

400

500

600

700

Wavelength/nm

Figure 8. (a) Absorption (black) and emission (red) spectra of 0.12 mM IQOOH in acetonitrile. The IQOOH ﬂuorescence quantum yield is Φ = 5.4·10^ꢀ4

.

Blue lines represent vertical excitation energies of IQOOH calculated using TD-B3LYP/6-31 + G^**level of theory (acetonitrile as a solvent using PCM

model). Oscillator strength axis corresponds to calculated transitions; (b) Difference density plots calculated using time-dependent DFT methods

(TD-B3LYP/6-31 + G**; PCM acetonitrile). Isovalue: (0.001). Red color represents depletion, while green color represents accumulation of the electronic

density in the excited state

wileyonlinelibrary.com/journal/poc

J. Phys. Org. Chem. 2013, 26 440–450

THERMOLYSIS AND PHOTOLYSIS OF ISOQUINOLINIUM HYDROPEROXIDE

not exhibit O – O bond dissociative character. However, these

calculations do not address whether the O – O bond dissocia-

tion occurs during the vibrational relaxation of the IQOOH S₁

state.

To investigate whether IQOOH photolysis occurs, the irradia-

tion of IQOOH at 400 nm was monitored using NMR, steady-state

UV/vis spectroscopy and fs TA spectroscopy. The NMR spectra of

photolyzed sample (Fig. 9) are similar to those collected during

RT thermolysis: the peaks of IQOOH decrease in intensity after

prolonged irradiation, with a concomitant growth of IQ amide

peaks at 7.8 and 8.4–8.8 ppm (Fig. 9a–d).

possible role of hydrogen bonds on the photophysics/photo-

chemistry of IQOOH.

The low quantum yield for IQOOH photolysis (Φ = 0.7% in

acetonitrile, Fig. 10a) and the low IQOOH ﬂuorescence yield

(Φ = 5.4·10^ꢀ4in acetonitrile) raise the question of what is the

process that causes deactivation of IQOOH and competes

efﬁciently with the O – O bond breaking and ﬂuorescence. To

investigate this in more detail, we collected the fs TA spectra of

IQOOH in acetonitrile (Fig. 11). Most of the transient signal

decays within the initial 50 picoseconds, conﬁrming that the

excited IQOOH indeed undergoes a fast deactivation process.

Three components are observed in TA spectra: (i) the initially

formed transient (red trace in Fig. 11a), which has a lifetime of

However, the photolysis of IQ amide promotes the appearance

of an additional product with singlets at 9.4 and 9.9 ppm, as well

as a quartet at 4.81 ppm. This product was identiﬁed as

isoquinolinium ion IQ⁺, whose NMR spectrum is shown in Fig. 9f.

The mechanism of IQ⁺formation is currently not known, but we

postulate that it originates due to the shift in the IQOOH/IQ⁺

equilibrium (IQOOH + H⁺! IQ⁺+ H₂O₂) due to the release of a

proton during the decomposition of IQO radical. We observed

similar behavior during the electrochemical oxidation of

pseudobases of ﬂavinium ion, which have the pseudobase pKa

value of 3.7.^[53]Since the IQOOH/IQ⁺conversion is expected to

occur at the pH = 5.03,^[69]the proton removal from IQO^.radical is

likely to shift the IQOOH/IQ⁺equilibrium and lead to the release

of IQ⁺. Thus, the photochemical decomposition of IQOOH is clearly

more complicated than the thermal process, presumably due to

higher energy provided by the 400 nm photons, allowing for more

possible reaction pathways.^[82]

The quantum yield for IQOOH photolysis was determined

using potassium ferrioxalate as a standard.^[58]The UV/vis absorp-

tion spectroscopy was used to monitor the change in IQOOH

concentration with irradiation time as the decrease of the 384

nm absorption band, while the formation of IQ amide was

monitored using the growth of the 360 nm band (Fig. (10a)).

The quantum yield for IQOOH photolysis in acetonitrile was

determined to be only 0.7%, suggesting that the excited-state

O – O bond breaking is not the major deactivation mechanism

of IQOOH S₁state. This ﬁnding was surprising given that some

sort of nonradiative process must be the cause for the very low

IQOOH ﬂuorescence quantum yield (0.05%). The photolysis is

even less efﬁcient in methanol (0.15%, Fig. 10b), implying a

t₁= 0.8 ps and consists of the ground-state bleach with l_max

=

389 nm and a weak broad excited-state absorption band that

covers the 400–800 nm region; (ii) the second transient with a

strong absorption at 441 nm (pink trace, Fig. 11a), which is

formed with a t₁= 0.8 ps lifetime and decays with t₂= 8.6 ps

lifetime. The decay of this transient is associated with a blue shift

of the absorption maximum, as can be seen by comparing the

t₁= 1 ps and t₂= 11 ps spectra; (iii) The third transient at 438

nm is very weak (yellow trace, Fig. 11a), and exhibits long

lifetime (remains unrecovered during the time window of our

experiment, which is 1.6 ns).

This fast excited-state decay is consistent with the low ﬂuores-

cence quantum yield of IQOOH. In addition, the ground-state

bleach at 389 nm almost completely recovers within 1 ns,

suggesting that no detectable photochemical process occurs,

at least not the one with the formation of long-lived products.

This is in turn consistent with the low photolysis quantum yield

of IQOOH. Based on these experimental ﬁndings, two possible

mechanisms for fast excited-state deactivation of IQOOH were

proposed: (i) the bond break/form model, in which the O–O bond

breaking from the S₁state of IQOOH is followed by a fast recom-

bination of IQO^.and OH^.radicals to regenerate the ground-state

IQOOH. In accordance with this model, the observed t₁=0.8 ps

lifetime can be assigned to the decay of IQOOH S₁state and

the rise of IQO^.radical signals due to O – O bond breakage

(IQOOH* ! IQO^.+ OH^.), while the t₂= 8.6 ps component can be

assigned to the recombination of the formed radicals to gener-

ate the S₀state of IQOOH (IQO^.+ OH^.! IQOOH). The weak

Figure 9. Changes in ¹H NMR spectra of 0.02 M IQOOH in CD₃CN illustrating the light induced decomposition after (a) 0 min; (b) 45 min; (c) 90 min and

(d) 150 min. The reference ¹H NMR spectra of IQ amide (0.02M in CD₃CN) and IQ⁺(0.05 M in CD₃CN) are presented in graph e and f

J. Phys. Org. Chem. 2013, 26 440–450

wileyonlinelibrary.com/journal/poc

R. KHATMULLIN ET AL.

1.5

1.0

0.5

0.0

1.5

1.0

0.5

0.0

a

b

_ACN=7

×

10^-3

_MeOH=1.5×

10^-3

250

300

350

400

250

300

350

400

450

Wavelength/nm

Figure 10. The changes in UV/vis absorption during 400 nm photolysis of 0.18 mM IQOOH after 0 (red), 5 (orange), 15 (green), 25 (maroon), 35 (cyan)

and 45 min (blue) of irradiation. The measurements were done in two solvents: (a) acetonitrile and (b) methanol. The pink dashed line in each graph

represents the UV/vis absorption of IQ amide

long-lived transient with lifetime t₃> 3 ns can be assigned to the

fraction of radicals that escape the recombination and continue

into the thermal reaction to generate IQ amide and water; (ii)

the S₁/S₀conical intersection model, according to which the S₁

state of IQOOH undergoes thermal S₁/S₀deactivation. According

to this model, the transient with t₁=0.8 ps lifetime can be

assigned to S_{1 !}S₀internal conversion, the t₂= 8.6 ps lifetime

can be assigned to the vibrational cooling of S₀state, while the

long-lived t₃> 3 ns component appears due to the slight red-shift

of the IQOOH ground-state absorption spectrum in a solvent that

has elevated temperature due to thermal equilibrium between

IQOOH and the solvent.

To experimentally distinguish between the break/form and

the conical intersection models, the photophysical behavior of

a model compound IQOH was investigated (Fig. 12). The lack

of O – O bond in IQOH restricts it from undergoing the break/

form deactivation mechanism. Interestingly, the TA spectra of

IQOH are surprisingly similar to those of IQOOH, with an absorp-

tion band at 434 nm and the bleach at 395 nm (Fig. 12a). The

signal dynamics exhibit three lifetimes (t₁= 0.8 ps and t₂= 10

ps, t₃> 3 ns), which are very similar to the lifetimes observed

in the case of IQOOH (Fig. 11). This large similarity between

IQOOH and IQOH photophysics strongly implies that the O–O

bond breaking is not involved in the fast excited-state deactiva-

tion of IQOOH. It was postulated that the S₁/S₀conical

intersection model is valid, and the following mechanism for

IQOOH photolysis is presented (Fig. 13): excitation of IQOOH at

400 nm generates “hot” S₁state with p,p* character. The short

t₁= 0.8 ps component is assigned to the S₁! S₀internal conver-

sion. This assignment is based on the fact that IQOH exhibits

stimulated emission at 545 nm, which decays with the similar

rate lifetime (t₁= 0.9 ps). The vibrationally excited S₀state further

undergoes intramolecular (IVR) and intermolecular (VET) vibra-

tional cooling with a t₂= 8.6 ps lifetime. The t₂= 8.6 ps lifetime

is consistent with the IVR and VET lifetimes reported using

IR-pump-UV/vis-probe experiments.^[83]The rate of IVR and VET

must be signiﬁcantly higher than the rate of O–O bond scission,

causing only less than 1% of vibrationally excited S₀population

to form IQ amide. The nuclear coordinate that leads to this fast

S₁! S₀internal conversion of both IQOOH and IQOH is currently

not known. It is interesting to note that we previously observed a

very similar fast S₁! S₀conversion in another hemiaminal,

named EtFlOH (N(5)-ethyl-4a-hydroxyﬂavin, Fig. S4, Supporting

information).^[84]In this study, we proposed that the excited-state

deactivation occurs due to S₁! S₀conical intersection reached

via sp³! sp²hybridization change of the N5 atom of the ﬂavin

moiety, followed by an out-of-plane distortion of the pyrimidine

ring. It seems possible that this fast excited-state deactivation is

general for other hemiaminal derivatives that are coupled with

the aromatic units.

100

50

100

50

a)

b)

50

0

50

427 nm

389 nm

430 nm

393 nm

0

-50

-100

-150

-50

-100

-150

-50

-100

-50

-100

0

20

40

60

350

500

600

350

500

600

0

20

40

60

Wavelength / nm

Time / ps

Wavelength / nm

Time / ps

Figure 12. (a) TA spectra of 1 mM solution of IQOH in acetonitrile after

the 400 nm excitation pulse at ꢀ1.1 (black), 0 (red), 0.2 (blue), 0.7 (dark

cyan), 4.0 (pink), 10 (cyan), 20 (orange), 200 (yellow) and 1309 ps

(green), respectively; (b) Decays dynamics of IQOOH transient absorption

signals collected at 439 and 393 nm, respectively. The black lines

correspond to ﬁts obtained using the following lifetimes: t₁= 0.8 ps,

t₂= 10 ps and t₃> 3 ns; (c) structure of IQOH

Figure 11. (a) TA spectra of 1 mM solution of IQOOH in acetonitrile after

the 400 nm excitation pulse at ꢀ4.9 (black), 0 (red), 0.2 (blue), 0.5 (dark

cyan), 1.0 (pink), 11 (cyan), 201 (yellow) and 301 ps (green) probe

delays; (b) Dynamics of IQOOH transient absorption collected at 427

and 389 nm. Black lines represent ﬁts obtained using the following

lifetimes: t₁= 0.8 ps, t₂= 8.6 ps and t₃> 3 ns

wileyonlinelibrary.com/journal/poc

J. Phys. Org. Chem. 2013, 26 440–450

THERMOLYSIS AND PHOTOLYSIS OF ISOQUINOLINIUM HYDROPEROXIDE

Last, we investigated whether the variations in solvent polar-

ity/viscosity can be used to manipulate the rate of nonradiative

S₁! S₀deactivation. Transient spectra of IQOOH in all solvents

Table 3. Lifetimes of components observed in transient

absorption measurements of IQOOH in different solvents

studied here are qualitatively similar (Fig. S5 and S6, Supporting

Information), and the signal dynamics (Fig. 14 and Table 3) are

not drastically affected by the solvent characteristics. In speciﬁc,

the polarity study shows that dynamics at 427 nm consist of two

components with lifetimes in the t₁=0.5–1.0 ps and t₂= 7.0–12.0

ps ranges, while the viscosity study shows the presence of only

one lifetime component that varies in the 3.5–4.5 ps range

(Table 3). It appears that the most dominant difference occurs

between the two sets of solvents; that is, when one switches

from aprotic solvents (benzene, chloroform, etc.) to the range

of protic solvents (methanol/ethylene-glycol mixtures). These

results suggest that the protic solvent-induced changes in

hydrogen bonding patterns of IQOOH cause the S₁! S₀conver-

sion to occur faster than in aprotic solvent. This increased rate of

deactivation in turn additionally reduces the quantum yield for

photolysis (from 0.7% in CH₃CN to 0.15% in methanol, Fig. 8).

Viscosity

Polarity

MeOH:EG Z/mPa t/ps Solvent

molar

fraction

Diel.

t₁/ps t₂/ps

ꢁs^[85]

constant^[86]

1:1

3.8 3.9 Benzene

2.31

9.08

37.5

1.0

1.0 12.0

0.8

0.5

9.0

0.3:0.7

0.15:0.85

0:1

7.4 3.5 CH₂Cl₂

12.0 3.7 CH₃CN

20.8 4.3 DMSO

8.0

7.0

46.68

Fs TA measurements reveal that the excited-state dynamics of

IQOOH are very similar to the model compound IQOH. Since IQOH

cannot undergo O – O bond breakage, it was proposed that the

surprising photostability of IQOOH could be ascribed to the fast

S₁/S₀thermal deactivation.

Acknowledgements

CONCLUSION

We thank Bowling Green State University and National Science

Foundation (CHE 1055397) for ﬁnancial support, Ohio Supercom-

puter Center for computational resources, Dr. Felix N. Castellano

for help with steady-state spectroscopic measurements. We give

special thanks to Dr. Robert Bach from University Of Delaware for

helpful discussions about TS calculations and mechanism of

IQOOH decomposition.

The RT thermolysis and photolysis of IQOOH were found to lead to

the formation of water and the corresponding amide, IQ amide.

The rate of thermal decomposition of IQOOH at RT is signiﬁcantly

higher than that of typical alkyl and aryl hydroperoxides, even

though the calculated bond dissociation energies are similar.

Surprisingly, this rate is greatly inhibited in the presence of small

amounts of methanol. Based on the TS structure calculated using

DFT, we propose that the RT thermolysis in IQOOH is facilitated

by the C1–C9 proton shift. In the presence of methanol, the proton

shift is less likely to occur, as the basicity of the nitrogen center is

reduced due to the hydrogen bonding with methanol. The photo-

chemical decomposition proceeds with a very low quantum yield.

REFERENCES

[1] D. Swern, in: Organic Peroxides, Wiley-Interscience, New York, 1970, 1.

[2] R. D. Mair, H. R. T. Hall, in: Organic Peroxides, Wiley-Interscience, New

York, 1972, 2.

[3] D. Cremer, in: The Chemistry of Functional Groups Peroxides, John Wiley

& Sons, New York, 1983.

[4] T. M. D. G. Hendry, L. Piszkiewicz, J. A. Howard, H. K. Eigenmann,

J. Phys. Chem. Ref. Data 1974, 3, 937.

S₁

NO

[5] Z. Cekovic, Tetrahedron 2003, 59, 8073.

[6] D. V. Avila, C. E. Brown, K. U. Ingold, J. Lusztyk, J. Am. Chem. Soc.

1993, 115, 466.

[7] C. Reichardt, J. Schroeder, D. Schwarzer, J. Phys. Chem. A 2007, 111, 10111.

[8] J. Wang, T. Tateno, H. Sakuragi, K. Tokumaru, J. Photochem. Photobiol. A

1995, 92, 53.

[9] B. Abel, J. Assmann, M. Buback, M. Kling, S. Schmatz, J. Schroeder,

Angew. Chem. Int. Ed. 2003, 42, 299.

N

NO

OOH

CI

400 nm

IQOOH

>1%

N

S₀

O

>99%

IQ Amide

[10] D. E. Falvey, B. S. Khambatta, G. B. Schuster, J. Phys. Chem. 1990, 94, 1056.

[11] K. U. Ingold, M. Smeu, G. A. DiLabio, J. Org. Chem. 2006, 71, 9906.

[12] J. Hartung, R. Stowasser, D. Vitt, G. Bringmann, Angew. Chem. Int. Ed.

1996, 35, 2820.

Figure 13. Proposed mechanism of IQOOH photolysis

[13] M. Jha, D. A. Pratt, Chem. Commun. 2008, 1252.

[14] Z. J. West, R. K. Adams, S. Zabarnick, Energy Fuel 2011, 25, 897.

[15] M. P. Richards, M. A. Dettmann, J. Agric. Food Chem. 2003, 51, 3886.

[16] J. P. Roozen, E. N. Frankel, J. E. Kinsella, Food Chem. 1994, 50, 39.

[17] D. C. Pascal, Macromolecules 2009, 42, 3469.

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

[18] K. U. Ingold, Chem. Rev. 1961, 61, 563.

[19] M. S. Kharasch, A. Fono, W. Nudenberg, J. Org. Chem. 1951, 16, 113.

[20] M. S. Kharasch, A. Fono, W. Nudenberg, J. Org. Chem. 1950, 15, 748.

[21] R. Hiatt, T. Mill, F. R. Mayo, J. Org. Chem. 1968, 33, 1416.

[22] R. Hiatt, T. Mill, K. C. Irwin, J. K. Castlema, J. Org. Chem. 1968, 33, 1421.

[23] R. Hiatt, T. Mill, K. C. Irwin, J. K. Castlema, J. Org. Chem. 1968, 33, 1428.

[24] R. Hiatt, K. C. Irwin, C. W. Gould, J. Org. Chem. 1968, 33, 1430.

[25] R. Hiatt, K. C. Irwin, J. Org. Chem. 1968, 33, 1436.

[26] J. L. Bolland, G. Gee, J. Chem. Soc. Faraday Trans. 1946, 42, 236.

[27] N. A. Milas, D. M. Surgenor, J. Am. Chem. Soc. 1946, 68, 205.

[28] V. Stannett, R. B. Mesrobian, J. Am. Chem. Soc. 1950, 72, 4125.

[29] K. U. Ingold, J. Phys. Chem. 1960, 64, 1636.

0

5

10 15 20 25

time / ps

0

10 20 30 40 50

time/ ps

Figure 14. Decay dynamics of 1 mM solution of IQOOH monitored at (a)

440 nm in ethylene glycol/methanol mixtures with relative viscosities: 3.8

(red); 7.4 (blue); 12 (dark yellow) and 20.8 (whine). (b) 427 nm in the polar

solvents: benzene (red); dichloromethane (blue); acetonitrile (dark yellow);

and dimethylsulfoxide (wine)

J. Phys. Org. Chem. 2013, 26 440–450

wileyonlinelibrary.com/journal/poc

R. KHATMULLIN ET AL.

[30] J. R. Thomas, J. Am. Chem. Soc. 1955, 77, 246.

[65] M. Cossi, V. Barone, R. Cammi, J. Tomasi, Chem. Phys. Lett. 1996,

255, 327.

[31] J. R. Thomas, O. L. Harle, J. Phys. Chem. 1959, 63, 1027.

[32] R. Hiatt, T. McCarrick, J. Am. Chem. Soc. 1975, 97, 5234.

[33] L. Bateman, H. Hughes, J. Chem. Soc. 1952, 4594.

[34] M. Makinen, A. Kamal-Eldin, A. M. Lampi, A. Hopia, J. Am. Oil Chem.

Soc. 2000, 77, 801.

[35] F. G. Gelalcha, Chem. Rev. 2007, 107, 3338.

[36] A. L. Baumstark, Bioorg. Chem. 1986, 14, 326.

[37] S. Ball, T. C. Bruice, J. Am. Chem. Soc. 1980, 102, 6498.

[38] P. C. B. Page, J. P. Heer, D. Bethell, B. A. Lund, Phosphorus Sulfur

1999, 153, 247.

[39] T. Nishio, J. Chem. Soc. Perkin Trans. 1991, 1, 1717.

[40] J. Rebek, R. Mccready, J. Am. Chem. Soc. 1980, 102, 5602.

[41] N. M. Kamerbeek, D. B. Janssen, W. J. H. van Berkel, M. W. Fraaije,

Adv. Synth. Catal. 2003, 345, 667.

[42] H. Schmaderer, P. Hilgers, R. Lechner, B. Konig, Adv. Synth. Catal.

2009, 351, 163.

[43] N. N. Schwartz, J. H. Blumbergs, J. Org. Chem. 1964, 29, 1976.

[44] N. Mora-Diez, S. Keller, J. R. Alvarez-Idaboy, Org. Biomol. Chem. 2009,

7, 3682.

[66] V. Barone, M. Cossi, J. Tomasi, J. Chem. Phys. 1997, 107, 3210.

[67] V. Barone, M. Cossi, J. Tomasi, J. Comput. Chem. 1998, 19, 404.

[68] M. J. T. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,

J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson,

H. Nakatsuji, M. Caricato, X. Li, H. P..Hratchian, A. F. Izmaylov,

J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota,

R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,

H. Nakai, T. Vreven, J. A. Montgomery, Jr, J. E. Peralta, F. Ogliaro,

M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov,

R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,

S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene,

J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts,

R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,

J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,

P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas,

J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 03, Gaussian,

Inc., Wallingford, CT, 2009.

[69] J. W. Bunting, W. G. Meathrel, Can. J. Chem. 1972, 50, 917.

[70] C. F. H. Tipper, J. Chem. Soc. 1953, 1675.

[71] J. R. Barker, S. W. Benson, D. M. Golden, Int. J. Chem. Kinet. 1977,

9, 31.

[45] A. L. Baumstark, P. C. Vasquez, J. Org. Chem. 1987, 52, 1939.

[46] A. L. Baumstark, P. C. Vasquez, J. Phys. Org. Chem. 1988, 1, 259.

[47] Y. Imada, H. Iida, S. Ono, S. I. Murahashi, J. Am. Chem.

Soc. 2003, 125, 2868.

[72] J. A. Kerr, Chem. Rev. 1966, 66, 465.

[48] P. Menova, F. Kafka, H. Dvorakova, S. Gunnoo, M. Sanda, R. Cibulka,

Adv. Synth. Catal. 2011, 353, 865.

[49] Y. Imada, H. Iida, T. Naota, J. Am. Chem. Soc. 2005, 127, 14544.

[50] A. B. E. Minidis, J. E. Backvall, Chem. Eur. J. 2001, 7, 297.

[51] Y. Imada, T. Naota, Chem. Rec. 2007, 7, 354.

[52] R. Lechner, S. Kummel, B. Konig, Photochem. Photobiol. Sci. 2010, 9, 1367.

[53] V. Sichula, Y. Hu, E. Mirzakulova, S. F. Manzer, S. Vyas, C. M. Hadad,

K. D. Glusac, J. Phys. Chem. B, 2010,114, 9452.

[73] R. D. Bach, O. Dmitrenko, J. Phys. Chem. B 2003, 107, 12851.

[74] M. Iwata, T. C. Bruice, H. L. Carrell, J. P. Glusker, J. Am. Chem. Soc.

1980, 102, 5036.

[75] S. I. Murahashi, T. Oda, Y. Masui, J. Am. Chem. Soc. 1989, 111, 5002.

[76] T. C. Bruice, J. Chem. Soc. Chem. Commun. 1983, 14.

[77] T. C. Bruice, J. B. Noar, S. S. Ball, U. V. Venkataram, J. Am. Chem. Soc.

1983, 105, 2452.

[78] T. C. Bruice, Acc. Chem. Res. 1980, 13, 256.

[54] E. Mirzakulova, R. Khatmullin, J. Walpita, T. Corrigan, N. M. Vargas-

Barbosa, S. Vyas, S. Oottikkal, S. F.Manzer, C. M. Hadad, K. D. Glusac,

Nat. Chem. 2012, 4, 794.

[79] U. V. Venkataram, T. C. Bruice, J. Chem. Soc., Chem. Commun. 1984,

899.

[80] G. da Silva, M. R. Hamdan, J. W. Bozzelli, J. Chem. Theory Comput.

2009, 5, 3185.

[81] S. K. Shin, S. O. Park, Y. S. Choi, H. L. Kim, C. R. Park, J. Phys. Chem. A

2001, 105, 10018.

[82] R. Schmidt, H. D. Brauer, J. Photochem. Photobiol., A 1986, 34, 1.

[83] J. Assmann, M. Kling, B. Abel, Angew. Chem. Int. Ed. 2003, 42, 2226.

[84] D. Zhou, E. Mirzakulova, R. Khatmullin, I. Schapiro, M. Olivucci,

K. D. Glusac, J. Phys. Chem. B, 2011,115, 7136.

[85] N. G. Tsierkezos, I. E. Molinou, J. Chem. Eng. Data 1999, 44, 955.

[86] F. Macdonald, Handbook of Chemistry and Physics 2008, 89th

edition, CRC Press.

[55] J. W. W. Bunting, Org. Prep. Proced. Int. 1972, 4, 9.

[56] E. D. Lorance, W. H. Kramer, I. R. Gould, J. Am. Chem. Soc. 2002, 124, 15225.

[57] M. Montalti, in: Handbook of photochemistry, CRC Press, LLC, 2006.

[58] C. G. Hatchard, C. A. Parker, Proc. R. Soc. London, Ser. A 1956, 235, 518.

[59] G. Li, K. D. Glusac, J. Phys. Chem. A 2008, 112, 4573.

[60] S. Yamaguchi, H. O. Hamaguchi, Appl. Spectrosc. 1995, 49, 1513.

[61] C. T. Lee, W. T. Yang, R. G. Parr, Phys. Rev. B: Condens. Matter 1988, 37, 785.

[62] A. D. Becke, J. Chem. Phys. 1993, 98, 5648.

[63] J. Tomasi, M. Persico, Chem. Rev. 1994, 94, 2027.

[64] M. Cossi, V. Barone, J. Chem. Phys. 1998, 109, 6246.

wileyonlinelibrary.com/journal/poc

J. Phys. Org. Chem. 2013, 26 440–450

Featured Compounds

1

2

Compound Details

Structure Search

Compound Details

Structure Search

FC-1

Article Doi

Thermolysis and photolysis of 2-ethyl-4-nitro-1(2H)-isoquinolinium hydroperoxide

DOI: 10.1002/poc.3107

Source and publish data:

Authors:

Article abstract of DOI:10.1002/poc.3107

Full text of DOI:10.1002/poc.3107

Products guided by the article

R&D Labs maybe for 1432028-34-6

Relevant to this article

Hot Product