Rh/TiO2-Photocatalyzed Acceptorless Dehydrogenation of N-Heterocycles upon Visible-Light Illumination

Article abstract of DOI:10.1021/acscatal.0c00556

TiO2 is an effective and extensively employed photocatalyst, but its practical use in visible-light-mediated organic synthesis is mainly hindered by its wide band gap energy. Herein, we have discovered that Rh-photodeposited TiO2 nanoparticles selectively dehydrogenate N-heterocyclic amines with the concomitant generation of molecular hydrogen gas in an inert atmosphere under visible light (λmax = 453 nm) illumination at room temperature. Initially, a visible-light-sensitive surface complex is formed between the N-heterocycle and TiO2. The acceptorless dehydrogenation of N-heterocycles is initiated by direct electron transfer from the HOMO energy level of the amine via the conduction band of TiO2 to the Rh nanoparticle. The reaction condition was optimized by examining different photodeposited noble metals on the surface of TiO2 and solvents, finding that Rh0 is the most efficient cocatalyst, and 2-propanol is the optimal solvent. Structurally diverse N-heterocycles such as tetrahydroquinolines, tetrahydroisoquinolines, indolines, and others bearing electron-deficient as well as electron-rich substituents underwent the dehydrogenation in good to excellent yields. The amount of released hydrogen gas evinces that only the N-heterocyclic amines are oxidized rather than the dispersant. This developed method demonstrates how UV-active TiO2 can be employed in visible-light-induced synthetic dehydrogenation of amines and simultaneous hydrogen storage applications.

Full text of DOI:10.1021/acscatal.0c00556

Subscriber access provided by University of Birmingham
Article
Rh/TiO2-Photocatalyzed Acceptorless Dehydrogenation
of N-heterocycles upon Visible-Light Illumination
Narmina O. Balayeva, Zamin Mamiyev, Ralf Dillert, Nan Zheng, and Detlef W. Bahnemann
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.0c00556 • Publication Date (Web): 09 Apr 2020
Downloaded from pubs.acs.org on April 9, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 44
ACS Catalysis
1
2
3
4
5
6
Rh/TiO2-Photocatalyzed Acceptorless Dehydrogenation of N-heterocycles upon Visible-
Light Illumination
7
8
9
Narmina O. Balayeva a*, Zamin Mamiyev b,c, Ralf Dillert a,c, Nan Zheng d, Detlef W.
Bahnemann a,c,e*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a Institute of Technical Chemistry, Gottfried Wilhelm Leibniz University of Hannover,
Callinstr. 5, D-30167 Hannover, Germany
b Institute of Solid State Physics, Gottfried Wilhelm Leibniz University of Hannover,
Appelstr. 2, D-30167 Hannover, Germany
c Laboratory of Nano and Quantum Engineering, Gottfried Wilhelm Leibniz University of
Hannover, Schneiderberg 39, D-30167 Hannover, Germany
d Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, AR
72701 (USA)
e Laboratory “Photoactive Nanocomposite Materials”, Saint-Petersburg State University,
Ulyanovskaya str. 1, Peterhof, 198504 Saint-Petersburg, Russia
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 44
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ABSTRACT:
TiO2 is an effective and extensively employed photocatalyst, but its practical use in visible-
light mediated organic synthesis is mainly hindered by its wide band gap energy. Herein
we have discovered that Rh-photodeposited TiO2 nanoparticles selectively
dehydrogenate N-heterocyclic amines with the concomitant generation of molecular
hydrogen gas in an inert atmosphere under visible light (λmax=453 nm) illumination at room
temperature. Initially, a visible-light-sensitive surface complex is formed between the N-
heterocycle and TiO2. The acceptorless dehydrogenation of N-heterocycles is initiated by
direct electron transfer from the HOMO energy level of the amine via the conduction band
ACS Paragon Plus Environment
2

Page 3 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
of TiO2 to the Rh nanoparticle. The reaction condition was optimized by examining different
photodeposited noble metals on the surface of TiO2 and solvents, finding that Rh0 is the
most efficient co-catalyst and 2-propanol is the optimal solvent. Structurally diverse N-
heterocycles such as tetrahydroquinolines, tetrahydroisoquinolines, indolines, and others
bearing electron-deficient as well as to electron-rich substituents underwent the
dehydrogenation in good to excellent yields. The amount of released hydrogen gas
evinces that only the N-heterocyclic amines are oxidized rather than the dispersant. This
developed method demonstrates how UV-active TiO2 can be employed in visible-light-
induced synthetic dehydrogenation of amines and simultaneous hydrogen storage
applications.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS: acceptorless dehydrogenation, N-heterocycles, visible-light, TiO2, surface
complex
INTRODUCTION:
Aromatic N-heterocycles are integral part of biologically active natural products that play an
essential role in medicine and pharmaceutical industries as reactive intermediates for synthetic
applications. Given the great importance of these compounds, sustainable methods that use
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 44
1
2
3
4
5
6
7
8
9
renewable and abundant sources, such as light, while avoiding the use of toxic and environmentally
unfriendly chemicals, are highly desirable. Among the synthetic approaches for aromatic N-
heterocycles, the dehydrogenation of saturated heterocycles is distinguished by high atom
economy, sustainability, and flexibility in the introduction of substituents at the targeted positions.
Although, the complete aromatization of N-heterocyclic amines is an endothermic reaction,
bearing the NH group in scaffold facilitates the dehydrogenation process in comparison with
1,2,3,4-tetrahydronaphtalenes.1,2,3 On the other hand, the complete dehydrogenation is much more
challenging than the partial dehydrogenation after the formation of C=N intermediates. For
example, due to the less positive oxidation potential of saturated 1,2,3,4-tetrahydroquinoline
(1,2,3,4-THQ) (+0.72 V vs. SCE) than of 1,2-dihydroquinoline (+0.91 V vs. SCE), the
dehydrogenation of the second hydrogen molecule is thermodynamically much more difficult than
the first one.4 Dehydrogenation of N-heterocycles can be classified into two main categories:
acceptor-assisted and acceptorless dehydrogenation. Although molecular oxygen is a much
greener acceptor than the traditionally used stoichiometric oxidants such as DDQ and sulfur,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-•
highly reactive oxygen species such as superoxide radicals (O2 ), hydroxyl radicals (OH•), and
hydrogen peroxide are usually produced in the former method. These species often lead to over-
oxidation of the product, resulting in poor selectivity and low yield of the product, as well as a
restriction of functional group tolerance. In contrast, an oxidant-free or acceptorless
dehydrogenation (ADH) method provides a more compelling and environmentally friendly
alternative for N-heterocyclic amine synthesis (Scheme 1). Since the liberation of hydrogen
molecules as the sole byproduct enables saturated N-heterocycles to function as potential
hydrogen-storage materials, this aspect enhances the appeal of developing the acceptorless
dehydrogenation of saturated N-heterocycles.
ACS Paragon Plus Environment
4

Page 5 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
The first homogenous catalyst system developed by Fujita et al. for the ADH of saturated N-
heterocycles was an iridium complex containing a pyridonate ligand.5 Subsequently, several
homogenous catalysts such as ruthenium-hydride complexes6, osmium7, iridium8–12, cobalt13,
iron3, and nickel14,15 complexes were reported for the acceptorless dehydrogenation of N-
heterocycles by several research groups. The transition metal complex could be substituted by a
base, KOtBu16, or a strong Lewis acid such as B(C6F5)317 that are able to form a frustrated Lewis
pair with saturated N-heterocycles. Regardless of the important progress made, most of the
reported ADH reactions in homogenous catalytic processes required high temperature (above
110°C), base/acid, or a toxic solvent such as trifluoroethanol. TEMPO-mediated electrolysis
enabled the acceptorless dehydrogenation to be performed at room temperature in an undivided
cell where dehydrogenation of N-heterocycles occurred at the anode with the release of H2 at the
cathode.18 However, the complexities in recyclability of the homogenous catalysts still remain as
the main concern. Therefore, heterogeneous catalysts were developed to improve the recyclability
of the catalyst. Diverse types of catalysts, which were typically composed of transition metals on
solid support, were reported. Some notable and recent examples included Pd-doped
hydrotalcites19, single cobalt atoms on an ordered porous nitrogen-doped carbon matrix20, cobalt
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
21
phenanthroline complex on exfoliated graphene oxide and core-shell iron on the exfoliated
graphitic oxide22. Similar to the homogenous alternatives, these methods were usually performed
at 110°C or above.
In fact, using visible-light enables both homogeneous and heterogeneous catalytic ADH of N-
heterocycles at room temperature in a better economic and environmentally friendly way. Usually,
a homogeneous photocatalytic ADH process was realized with dual catalysis merging photoredox
catalysis with transition metal catalysis. Y. Li et al. reported the first ADH of cyclic amines at
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 44
1
2
3
4
5
6
7
8
9
ambient temperature by using homogeneous ruthenium complex (Ru(bpy)3Cl2·6H2O) and
Co(dmgH)2PyCl photocatalysts in ethanol (EtOH).4 Furthermore, the same dual catalysis system
was adopted for the acceptorless dehydrogenative coupling reaction of saturated N-heterocycles.23
Modification of the Co complex allowed the ADH of N-heterocycles to be performed in water.24
Kanai et al. developed a different dual catalyst system, which was composed of an acridinium
photoredox catalyst and a palladium catalyst to catalyze the ADH of N-heterocycles.25 Most
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
26
recently, a single catalyst system such as a binuclear platinum(II) diphosphite complex and a
supramolecular assembly consisted of a carbazolyl dicyanobenzene fluorophore with four
cobaloximes was reported for the ADH of N-heterocycles.27 Even though these reported methods
performed at room temperature with the assistance of blue LED light, the recyclability of these
photocatalysts is still challenging. Recently, ADH of N-heterocycles has been reported using
29
heterogeneous h-BCN 28 and MoS2/ZnIn2S4 photocatalysts. However, in the former method the
dehydrogenation of the limited number of substrates was possible at a comparable higher photon
energy (λmax=420 nm) and in the latter system the catalyst was effective only for partial
dehydrogenation of 1,2,3,4-tetrahydroisoquinoline to 3,4-dihydroisoquinoline. To date, there are
only these two reports about the ADH of N-heterocycles employing visible-light active
heterogeneous photocatalysts.
In the present work, we have targeted the titanium dioxide (TiO2) based catalytic system for the
ADH of N-heterocycles. TiO2 is a relatively low cost, earth-abundant, and commercially available
semiconducting material with large bandgap energy (3.2 eV for pristine anatase) and appropriate
redox potentials. Together with several superior characteristics such as non-toxicity, biological,
chemical, and thermal stability, it excels from other semiconductors. Therefore, it has been broadly
used in many applications, including energy (water splitting),30–32 building and household
ACS Paragon Plus Environment
6

Page 7 of 44
ACS Catalysis
1
2
3
4
5
6
materials (bleaching, anti-fogging, toothpaste, paints, sunscreens),33,34 and environmental (air and
water purification)35,36. Moreover, considerable efforts have been devoted to fundamental
7
8
9
37–40
transformation and chemical reactions in organic synthesis over TiO2 photocatalysts
succeeding the pioneer contributions of Frank, Reiche and Bard.41,42 However, two main
drawbacks of TiO2, large bandgap energy and fast charge carrier recombination, severely limit its
use as a visible-light photocatalyst in organic synthesis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The inherent drawbacks of UV-light active TiO2 could be surmounted by coordinating it with
several UV-light active adsorbates (ligands), forming a visible-light responsive surface complex.
Upon visible light excitation, the ground state electrons of the surface adsorbate are directly
transferred into the conduction band of TiO2 without involving the excited state, which is referred
to as a ligand-to-metal charge transfer (LMCT) process.43 Lang et al. reported the visible-light-
driven oxidation of primary and secondary amines into the corresponding imines on TiO2
employing several ligands such as catechol44, phenol45, salicylic acid46 and N-
hydroxyphthalimide47. However, appropriate electron donors are always needed in these systems
to regenerate the employed ligands. Our group has recently reported the selective, visible-light-
driven oxidative dehydrogenation of N-heterocyclic amines on the TiO2 surface, where amines act
as ligands and do not require donors due to their self-oxidation.48 Nevertheless, 4-amino-TEMPO
was involved in enhancing the reaction selectivity, which was hindered by the photogenerated
reactive oxygen species.
In our continuous efforts to investigate the selective oxidation of N-heterocycles over TiO2, we
envisaged the possibility of the ADH reaction of tetrahydroquinolines, tetrahydroisoquinolines,
indolines, and several other N-heterocycles using noble metal photodeposited TiO2 nanoparticles
in an inert atmosphere. Thus, amines are adsorbed at the TiO2 surface resulting in a surface
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 44
1
2
3
4
5
6
7
8
9
complex suitable to absorb visible light. After light absorption, the excited surface complex injects
an electron in the conduction band of the TiO2 photocatalyst. This conduction band electron can
be transferred to a suitable acceptor being adsorbed at the photocatalyst particle. However, it is
known that the transfer of an electron to H+ requires a co-catalyst, such as a noble metal.33
Therefore, we have combined two separate strategies for achieving ADH of N-heterocycles: (i)
employ noble metal nanoparticles photodeposited on TiO2 to retard the charge carrier
recombination while facilitating H2 evolution; (ii) use an N-heterocyclic amine-TiO2 surface
complex for visible light activation. This methodology suggests that surface adsorbed
tetrahydroquinolines, tetrahydroisoquinolines, indolines, and other N-heterocycles can be easily
dehydrogenated into the corresponding imines under visible-light illumination. The high
chemoselectivity of the overall system could be achieved due to the intrinsic lack of TiO2 to
respond to visible-light, which prevents the hole generation in the valence band. Moreover, using
an inert atmosphere avoids the photogeneration of the reactive oxygen species during the
dehydrogenation reaction.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thus, visible-light-mediated ADH of tetrahydroquinolines, tetrahydroisoquinolines, indolines,
imidazoline and other N-heterocycles is successfully facilitated in good to excellent yields using
Rh/TiO2 photocatalyst. Molecular hydrogen is produced as the sole byproduct in a yield that
usually matches that of the aromatized heterocycle.
Herein, we report the first example of acceptorless dehydrogenation of saturated N-heterocycles
by TiO2 under visible light at room temperature. The approach of using both surface-bound ligands
and surface-deposited nanoparticles to fine-tune catalytic activities of TiO2 will have broad
implications in the development of TiO2 and other wide band gap semiconductors as the
photocatalyst for visible-light-driven organic synthesis.
ACS Paragon Plus Environment
8

Page 9 of 44
ACS Catalysis
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
Photodeposition of noble metals on the surface of the TiO2:
8
9
The catalysts for the dehydrogenation of saturated N-heterocyclic amines were synthesized with
the simple photodeposition methods. In brief, the salts of noble metals with the appropriate amount
(0.1-5 wt%) with respect to TiO2 were dissolved in 30 mL of a mixture of 30 vol. % methanol and
deionized water. Additionally, 3 g of the commercially available Hombikat TiO2 UV-100
(Venator) with 100 % anatase crystal phase and ~320 m2/g specific surface area was added to the
solution inside the photoreactor. Subsequently, the photoreactor (50 mL) was tightly closed,
purged with Ar for 20 min and irradiated by UV-A light with the intensity of 1 mW/cm2 with
vigorously stirring for 24h at ambient temperature. In the next step, the obtained suspension was
rinsed three times with deionized water in conical bottom plastic vials and centrifuged each time
to eliminate irrelevant anions from the surface. Finally, the obtained residue was dried in a muffle
oven at 110◦C and ground to a fine powder properly.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Characterizations:
The crystalline phase and lattice parameters of commercial TiO2 (UV-100) and other surface-
modified TiO2 with noble metal nanoparticles were studied by X-ray diffractometer (XRD) using
Bruker D8 Advance apparatus at room temperature employing a Cu Kα radiation source (λ=1.5406
Å) over a 2θ range from 10° to 120°, with a 0.05° step width (SI, Figure S4). The morphologies of
the samples were investigated by transition electron microscopy (TEM) on a Tecnai G2 F20 TMP
device (FEI Company) operated under an acceleration potential of 200 kV with a FEG field (SI,
Figure S5). Brunauer–Emmett–Teller (BET) measurements were performed to determine the
specific surface area of the photocatalysts by a single-point through physisorption of nitrogen at
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 44
1
2
3
4
5
6
7
8
9
the liquid nitrogen temperature (−196 °C) using a FlowSorb II 2300 device from Micromeritics
Instrument Company (SI Table S1). The diffuse reflectance UV-vis spectra (DRS) of the
photocatalysts were recorded at room temperature via Varian Spectrophotometer Cary-100 Bio
from Agilent technologies for the optical properties of the catalysts. White BaSO4 powder was
used as a baseline reference material which has absolute reflectance (100%). All reflectance
spectra were converted into F(R) by applying the Kubelka–Munk function to observe the shift of
the spectra due to different metal loading on the TiO2 surface. X-ray photoelectron spectroscopy
(XPS) experiments were performed for all metal-photodeposited TiO2 samples to study the
oxidation states of the metal islands (SI, Figure S7). A Leybold Heraeus XPS device, equipped
with an X-ray Al Kα beam as an excitation source and the binding energy corresponding to the
metals determined with a calibrated 1s Carbon peak. XPS 4.1 software was used to fit the curves
of raw data by the Gaussian/Lorentzian fitting equation. The content of metal nanoparticles on the
TiO2 surface was determined using ICP-OES measurement. Instrumental conditions and detailed
characterization techniques such as XRD, TEM, XPS, BET, Tauc plot of UV-vis spectra (SI,
Figure S6), the experimental procedure of organic synthesis, and GC-TCD calibration spectra of
molecular hydrogen gas (SI, Figure S3) are presented in the Supporting Information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Photocatalytic reaction procedure:
An oven-dried 10 mL reaction vial was used as a reactor. 5 mg photocatalyst was dispersed in 2
mL appropriate solvent, and then 0.2 mmol reactant (various N-heterocyclic amines) was added.
After sealing the reactor tightly, the reaction mixture was sonicated for 10 min in order to disperse
powder particles properly. Then, the suspension was purged for 15 min with argon gas to exchange
all air inside the reactor, and the inert atmosphere (without N2 and O2) was established by analyzing
the headspace of the reactor by gas chromatography (Shimadzu 8A, TCD detector). Depending on
ACS Paragon Plus Environment
10

Page 11 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
the substrates, the photocatalytic experiments were carried out for 24 h or longer under visible-
light illumination (LED array with λmax=453 nm) with continuous shaking. Afterward, the evolved
hydrogen gas was analyzed by GC-TCD. Later on, the catalyst was separated from the reaction
mixture by centrifuging, and the product was isolated by flash column chromatography. The purity
of the isolated compound was confirmed by both 1H NMR and 13C NMR measurements recorded
in CDCl3. In the case of optimization of the reaction condition, shown in Table 1, benzyl alcohol
(10.3 µL, 0.1 mmol, 50 mol%) was added to the solution as an internal standard. 1,2,3,4-
tetrahydroquinoline was chosen as the representative substrate to optimize the reaction conditions
(SI, Figure S2). The photonic efficiency (ζ) of the different noble metal photodeposited TiO2 was
calculated on the basis of the results shown in Table 1 (See SI for the calculation procedure).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION:
As mentioned in the introduction, a prerequisite for the desired ADH reaction of N-heterocyclic
amines is the adsorption of the reactant at the photocatalyst surface, yielding a visible-light-
absorbing surface complex. We have measured the UV-Vis spectra of alcoholic suspensions
containing an organic substrate (1,2,3,4-tetrahydroquinoline, 1,2,3,4-tetrahydroisoquinoline, or
indoline) and the metal-loaded photocatalyst 1 wt% Rh/UV-100 (Figure 1a). A significant redshift
became obvious from Figure 1a, indicating the formation of the desired surface complex. Initially,
we intended to perform an ADH reaction of 1,2,3,4-tetrahydroquinoline (1,2,3,4-THQ) 1a as a
model substrate under visible-light (λmax = 453 nm) illumination at room temperature in an inert
atmosphere over bare TiO2. As expected, bare TiO2 was found to be inactive in the
dehydrogenation reaction of 1,2,3,4-THQ (entry 1 in Table 1) due to the hindered electron transfer
from the conduction band of TiO2 to the H+. To enhance the necessary reduction of protons, TiO2
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 44
1
2
3
4
5
6
7
8
9
loaded with Cu2+, Au0, RuO2, Pd0, Pt0, and Rh0 nanoparticles have been employed as the
photocatalysts. Results are presented as entries 2-7 in Table 1. It becomes obvious from these data
that the highest yields at constant illumination time are obtained with Pd0, Pt0, and Rh0. The order
of activity of the metals was found to be Rh ˃ Pt ˃ Pd. RuO2 and Au photodeposited TiO2 showed
dramatically low photocatalytic activity for the ADH reaction of 1,2,3,4-THQ. Interestingly, UV-
vis diffuse reflectance spectra show that only gold and ruthenium oxide nanoparticles
photodeposited TiO2 has a strong plasmonic absorption peak at around 400-630 nm. While Pt and
Pd photodeposited TiO2 exhibit the weak absorption edge at around 600-800 nm, the rhodium
photodeposited TiO2 absorb only has in the UV region (Figure 1b). The visible-light absorption
of ruthenium oxide and gold co-catalysts possible hinter the dehydrogenation reaction due to other
competitive visible-light-induced reaction pathways. These results were consistent with our
suggestion that the dehydrogenation was initialized by visible-light excitation of the amine-TiO2
complex. After revealing the crucial role of different co-catalysts in the system, experimental runs
were performed with Rh/TiO2 and 1,2,3,4-THQ in other solvents. The ADH of 1,2,3,4-THQ in
acetonitrile resulted in product formation with moderate yields (Table 1, entry 8). In methanol and
ethanol, only small amounts of the desired product were formed (Table 1, entries 9 and 10). No
reaction was observed during shaking of 1,2,3,4-THQ in 2-propanol in the presence of Rh-loaded
TiO2 in the dark (Table 1, entry 12). Visible-light illumination of Rh-loaded TiO2 and 2-propanol
suspension, containing no 1,2,3,4-THQ, showed no formation of H2, thus evincing that the alcohol
is not oxidized under these experimental conditions (Table 1, entry 11). Visible-light illumination
of the solution of 1,2,3,4-THQ in 2-propanol containing no photocatalyst showed no
dehydrogenation of the reactant (Table 1, entry 16). These results reveal the importance of TiO2
and visible-light illumination in the ADH reaction. On the other hand, Pd/C and other transition
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
12

Page 13 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
metals ion (iron, nickel, niobium) grafted on TiO2 did not show any photocatalytic activity for the
ADH reaction of 1,2,3,4-THQ, except 1 wt% Cu (II) grafted TiO2 (Table 1, entry 2) which is still
too low in comparison with the active noble metal nanoparticles. Although Pd/C has been reported
as a selective thermal catalyst for dehydrogenation of N-heterocycles, high reaction temperatures
(T≥140°C) have always been required.49 However, the visible light dehydrogenation of the N-
heterocycles occurs through the surface complexation of the amine and the noble metal
photodeposited TiO2 at room temperature in which the Pd/C catalyst was inactive. Hence, the
presence of TiO2 catalyst as a semiconductor is a prerequisite for visible light activation. As
comparison to the photodeposition of noble metals, the transition metals ion (iron, nickel, niobium)
grafting on TiO2 showed almost inactivity, except for Cu(II). The difference in reactivity could
be explained by the fact that, because the noble metal islands act as electron sinks to induce a
cathodic shift of the quasi-Fermi level, the contact between the metal and the semiconductor
facilitates the reduction of protons. On the other hand, this process was not expected to occur
between the TiO2 semiconductor and the metal ions. In addition, transition metal ions, such as
nickel, iron, and niobium, are not as good an electron donor as noble metal, therefore, rapid
recombination occurs before the reduction of the proton.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to further increase the efficiency of photocatalytic reactions, we also investigated the
optimal photocatalyst loading (1 mg, 2 mg, and 5 mg 1 wt% Rh/TiO2) for the ADH reaction of
1,2,3,4-THQ (Figure 2). A significant decrease in the photocatalytic performance was observed
when the amount of 1 wt% Rh/TiO2 photocatalyst in a 1,2,3,4-THQ-containing suspension was
reduced from 2.5 g/L (5 mg of the photocatalyst) to 0.5 g/L (1 mg of the photocatalyst).
Furthermore, the time-dependent experiments with varying the amount of the photocatalyst in the
reaction system were performed, revealing that at the initial step, the reaction kinetics was faster
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 44
1
2
3
4
5
6
7
8
9
independent from the catalyst loading amount. Nevertheless, it was observed that the ADH
reaction of 1,2,3,4-THQ in the presence of 1 mg of 1 wt% Rh/TiO2 photocatalyst almost stopped
at a quinoline yield of ~40% after 15 h of illumination. Extending the illumination time to 24 h did
not result in a significant increase of the amount of product. This could be attributed to the
poisoning of catalytic active sites by the adsorption of product (quinoline) molecules. Thus, upon
visible light illumination, the amine molecules adsorbed on the surface of TiO2 were
dehydrogenated and vacated from the surface. Meanwhile, some final oxidation product
(quinoline) molecules was continuously accumulated on the surface of TiO2 thus competing
against the adsorption of amine molecules at TiO2 active sites. A higher photocatalyst loading
amount up to 5mg (2.5 g/L) showed a higher reactivity within a shorter reaction time and the amine
was completely converted to the product (quinoline) after 24 h of illumination due to the presence
of an appropriate number of active sites on the photocatalyst. Therefore, suitable amount of the
catalyst and the reaction time was established to be 5 mg and 24 h for the ADH of 0.2 mmol
1,2,3,4-THQ, respectively. Hence, all the following experiments have been performed in i-PrOH
with 5 mg Rh/TiO2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Moreover, a variation of Rh nanoparticles content was also investigated. The experimental results
are presented in Figure 3. It was observed that the Rh0 loading between 0.5 wt% and 1 wt% gave
an excellent yield within 18 h reaction time. However, 5 wt% Rh0 co-catalyst deposited TiO2
showed significantly poor photocatalytic performance. This results could be due to the fact that
the increase of the content of Rh metal nanoparticles up to 5 wt% with respect to TiO2 leads to
aggregation of Rh metals particles on the TiO2 surface thus lowering the surface area necessary
for the interaction with protons. The Rh nanoparticles could consequently act as recombination
centers rather than the electrocatalysts for the hydrogen evolution reaction accompanying the ADH
ACS Paragon Plus Environment
14

Page 15 of 44
ACS Catalysis
1
2
3
4
5
6
reaction. Additionally, it might occupy the Lewis acidic sites on the TiO2 surface thus hindering
the surface complexation of amine molecules.
7
8
9
The ratio between the Rh and TiO2 nanoparticles for the 0.1 wt%, 1 wt%, and 5 wt% Rh
photodeposited TiO2 samples was calculated to determine the surface coverage of the employed
photocatalysts (See SI for the calculation procedure).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The TEM images of 1wt% Rh/TiO2 sample clearly showed that the surface of TiO2 was not
homogenously covered with Rh nanoparticles (Figure S5). Metallic Rh nanoparticles were in
contact with the semiconductor nanoparticles and thus were attached to the TiO2 surface. By
calculating the ratio between 3nm Rh particles and 5nm TiO2 particles obtained from TEM images,
it was found that one Rh nanoparticle exists in every ~70 TiO2 nanoparticles, thus the surface
coverage of 1 wt% Rh nanoparticles decorated on TiO2 nanoparticles is very low. Therefore, the
light absorption properties of the TiO2 particles was not changed significantly in 1 wt% Rh/TiO2
powder.
The range of substrates for different N-heterocyclic amines (0.2 mmol) was explored at room
temperature in an inert argon atmosphere after optimization of the reaction condition using 5 mg
1 wt% of Rh photodeposited TiO2 as a photocatalyst and 2 mL of i-PrOH as a solvent.
Under these optimized experimental conditions using 5 mg 1 wt% Rh/TiO2 photocatalyst in 2
mL i-PrOH as the solvent, 1,2,3,4-THQ (1a) was converted with a photonic efficiency of 7.6 %
giving quinoline (2a) and molecular hydrogen in 96% and 99% yields, respectively. The high
yields prompted us to investigate the ADH reaction of eight different 1,2,3,4-
tetrahydroquinolines substituted with electron-donating, electron-withdrawing, and alkyl groups
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 44
1
2
3
4
5
6
in different positions of the benzofused N-heterocycle. Most reactants were found to yield the
corresponding quinolines in good to excellent yields (54% - 93%, see Table 2). Exceptions were
7
8
9
7-nitro-1,2,3,4-THQ (1b) and 7-hydroxy-1,2,3,4-THQ (1c) reactants, that afforded no or
very little amount of the products. The reactant bearing the strongly electron-withdrawing
nitro group (1b) was recovered almost quantitatively after 24 h exposure in the presence
of Rh/TiO2 in i-PrOH. Neither the desired 7-nitroquinoline (2b) nor molecular hydrogen
were produced in detectable amounts. The analogous reaction of the reactant 1c having
a strongly electron-donating OH group at C-7 gave the products 7-hydroxyquinoline (2c)
and molecular hydrogen with very low yields of only 7%. This phenomenon was also observed
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
in our previous investigation, where the results suggested that the hydroxyl group and the amino
group were likely to compete for adsorption sites on the titanium surface. Thus, protecting the
hydroxyl group as a TBS ether removed the complexation of the OH group with TiO2, and indeed,
the oxidative dehydrogenation of 7-TBSO-THQ was afforded in a high yield.48
The light-induced ADH reactions of 1,2,3,4-THQs with substituents at C-6, i.e., 6-methoxy-
1,2,3,4-THQ (1d), 6-chloro-1,2,3,4-THQ (1e) and 6-methyl-1,2,3,4-THQ (1f), resulted in the
quinoline formation with yields between 68% and 90%. Molecular hydrogen was generated in
these reactions in the amounts to be expected on the basis of the stoichiometry. However,
a longer reaction time (36 h) was required for 1e and 1f to furnish the products in ca. 90%
ACS Paragon Plus Environment
16

Page 17 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
yields. A moderate product yield of 54% was determined for the ADH reaction of 2-methyl-
1,2,3,4-THQ (1g) after 24 h visible light illumination. Initially, we thought that the steric
effect exhibited by the methyl group might hinder the interaction between the NH group
of the reactant and the TiO2 surface, which is necessary for the absorption of the visible
light. However, the disubstituted 8-carboxyl-2-methyl-1,2,3,4-THQ (1h) was dehydrogenated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
yielding the 8-carboxyl-2-methyl-quinoline (2h) with 93% yield within 36 h of illumination.
Hence, the difference in the product yield between 1g and 1h suggested that the steric effect
displayed by the substituents at C-2 and C-8 had minimal effect on the reaction. The successful
ADH reaction of 4-methylthio-2-phenyl-1,2,3,4-THQ (1i) also indicated a lack of a significant
steric effect by the substituent at C-2. Despite the presence of a bulkier phenyl group at
C-2, the product 2i was formed within 36 h in a good yield of 61%. However, because the
limited number of employed reactants were examined, it would be difficult to conclusively
assess the influence of steric and electronic effects on the reactivity of substituted 1,2,3,4-
tetrahydroquinolines.
The optimal reaction condition was extended to other types of saturated N-heterocycles (1j – 1m).
It was observed that the tetrahydroisoquinolines in ADH reaction behaved differently compared
to tetrahydroquinolines where partial dehydrogenation products (3,4-dihydroisoquinolines) were
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 44
1
2
3
4
5
6
7
8
9
obtained along with complete dehydrogenation products (isoquinolines) after 48 h visible-light
illumination (Table 3). Thus, 1,2,3,4-tetrahydroiso-quinoline without any substituent group
afforded the complete dehydrogenation product, isoquinoline (2j) in 48 % yield with the partial
dehydrogenation product, 3,4-dihydroquinoline (2j*) in 39% yield. Tetrahydroisoquinolines
bearing 1-methyl (1k) or 1-phenyl (1l) groups afforded mainly the partial dehydrogenation product
with the complete dehydrogenation product as the minor product. However,
tetrahydroisoquinoline bearing two methoxy groups at C6 and C7 positions (1m) underwent
complete dehydrogenation in excellent yield (2m, 88 %). The high reactivity was likely due to
strong electron-donating methoxy groups that made the oxidation facile.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
When the optimal reaction conditions (5 mg 1wt% Rh/TiO2 in 2 mL i-PrOH, 0.2 mmol reactant,
Ar athmosphere, 453 nm illumination) were applied to indolines, the unsubstituted indoline (1n)
as well as the 2-methylindoline (1o) were converted to the indoles in 73% and 49% yield,
respectively (Table 4). Interestingly, the ADH of 6-(trifluoromethyl)indoline (1p), gave the
product 2p in excellent yield (86 %). It seems that the strong electron-withdrawing group did not
hinder the dehydrogenation reaction.
Decarboxylation was observed with the indoline substrates bearing a 2-carboxyl group in the C2
position (1q and 1r). In both examples, indole was the isolated product. However, when the
carboxyl group was protected as a methyl ester, the decarboxylation was prevented entirely, and
methyl 1H-indole-2-carboxylate (2s) was obtained in 75% yield. It has been reported that a
chemical bond is formed between the carboxylic group and the Brønsted acid site at the TiO2
surface (Ti-OH).50 FTIR study revealed that the interaction of the carboxyl group with the TiO2
surface is more favorable than the interaction of the amine group with the oxide surface.51 On the
other hand, it is well-established that Photo-Kolbe decarboxylation occured when acetic acid
ACS Paragon Plus Environment
18

Page 19 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
adsorbed on the surface of TiO2, which has been irradiated with UV light.52 Although visible-light
is employed in our system, the reaction exhibited that decarboxylation still was favorable in the
C2 position of indoline before the dehydrogenation reaction. Recently, visible-light
decarboxylation of the same substrate 1q was studied on metal-halide perovskite in N2 atmosphere,
however both decarboxylation and dehydrogenation occurred in an O2 atmosphere.53 According
to the previous works, we have proposed a decarboxylation and dehydrogenation reaction
mechanism of indoline-2-carboxylic acid on the Rh/TiO2 surface (Scheme 2). We suppose that the
carboxyl group interacts with Ti-OH Brønsted acid site while the amine group forms a surface
complex with Ti+ Lewis acid site of TiO2. The latter complex, as the visible-light-absorbing
species, initiates the photon absorption and amine cation radical formation by the transfer of a
single electron to the conduction band of TiO2. Subsequently, the substrate undergoes the
decarboxylation through C-C bond cleavage, and an alpha-amino radical is generated. The radical
is oxidized yielding a cation that loses a proton to form a double bond and has thus been
dehydrogenated.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In addition, the same optimized reaction condition employed for the ADH of THQs was applied
for the ADH of N-heterocycles containing two nitrogen atoms (Scheme 3). 2-Phenyl-1H-
imidazoline (1t), a different class of N-heterocycles bearing two nitrogen atoms, underwent the
ADH reaction uneventfully, producing product (2t) in excellent yield (94 %). However, 1,2,3,4-
tetrahydroquinoxaline (1u) failed to participate in the ADH, although it underwent the oxidative
dehydrogenation with molecular oxygen in our previous study.48
After establishing the scope of the substrates, we investigated the minimal structure requirement
for the ADH (Scheme 3). Initially, completely saturated perhydroquinoline was subjected to
optimal conditions. This substrate is also attractive from the viewpoint of H2 storage materials as
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 44
1
2
3
4
5
6
7
8
9
it could theoretically release five molecules of hydrogen. However, only 96 μmol H2 was released
along with a poor conversion to 5,6,7,8-tetrahydroquinoline (2v*). This result indicated that having
an aromatic ring fused to the saturated N-heterocycle was essential for the complete
dehydrogenation. Surprisingly, the substrate 8-acetyl-5,6,7,8-tetrahydroquinoline (1s), which does
not have an NH group, furnished the completely aromatized product (2w) in 62 % yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
When the amine group was protected by a Boc group as shown in (1x), the surface complexation
with titania seemed to be completely inhibited, since the targeted product (2x) was obtained in
only 5% yield with H2 in 6% yield. This data also supported that formation of the surface complex
between the carboxyl group and titania, is not responsible for the visible-light absorption. These
studies suggested that having a Lewis basic nitrogen atom in conjunction with a fused arene or N-
heteroarene was critical for a successful acceptorless dehydrogenation reaction.
Since we have discussed that amine group adsorbed onto TiO2 surface initiates the visible-light
absorption and electron-transfer reactions, it was unexpected to be observed the ADH of 8-acetyl-
5,6,7,8-tetrahydroquinoline (1w) which is not bearing an amine group. Therefore, the origin of the
light-absorbing-species was further investigated by UV-vis spectroscopy. The bright yellow
substrate 1w showed absorption up to 475 nm. Moreover, adding 1 wt% Rh/UV-100 catalyst into
the solution of 8-acetyl-5,6,7,8-tetrahydroquinoline in i-PrOH demonstrated extended optical
absorption up to 525 nm (Figure 4). These spectroscopic analyses evince that a prerequisite for
the visible-light-mediated photocatalytic ADH is either substrate, which absorbs visible-light or a
surface complex due to surface interaction between 8-acetyl-5,6,7,8-tetrahydroquinoline and TiO2.
Therefore, we assumed that 8-acetyl-5,6,7,8-tetrahydroquinoline acts as the antenna molecule and
initiate a visible-light ADH reaction (Scheme 4). Hence, initially, upon visible-light illumination,
ACS Paragon Plus Environment
20

Page 21 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
an electron is excited from the HOMO to the LUMO energy level of the organic reactant. In the
next step, the excited substrate transfers its electron to the conduction band of the semiconducting
TiO2. The electron is consecutively transferred to the noble metal attached to the TiO2 surface
where it becomes temporarily stored. The formed cation of the substrate generated by electron
transfer is a strong Brønsted acid, which undergoes the first deprotonation and subsequently forms
the partial dehydrogenation product. Since the partial dehydrogenation product was not observed
after the photocatalytic experiment, it is assumed that the second dehydrogenation occurs
following the same reaction pathway to finally yield the 8-acetyl-quinoline (2w). Control studies
revealed that no transformation of 8-acetyl-5,6,7,8-THQ was achieved either in the absence of 1
wt% Rh/TiO2 or visible-light thus evincing the importance of the Rh/TiO2 photocatalyst and the
visible light in the ADH reaction.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
This result underlines that aromatization of cyclohexane-fused pyridine bearing acetyl group in
the presence of 1 wt% Rh/TiO2 in an inert atmosphere upon visible-light illumination could be
realized. However, the conclusion requires an in-depth investigation of the broad range of
substrates, which opens a new pathway in the visible-light-mediated ADH of cyclohexane-fused
pyridines involving TiO2 photocatalyst. A better understanding of the underlying reaction
mechanism and optimization of the reaction condition might improve the moderate yield up to
excellent yield.
Furthermore, the reusability of 1 wt% Rh/TiO2 was examined by carrying out six successive
experimental cycles with 1,2,3,4-THQ as the reactant in order to demonstrate the advantage of
using heterogeneous photocatalyst. The catalyst was collected by simple centrifugation and was
reused without further treatment until the 4th cycle (See SI for the recycling procedure). Although
the selectivity of the ADH reaction was found to be stable (>96 %), the yield of the product
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 44
1
2
3
4
5
6
7
8
9
gradually decreased after four experimental runs (Figure 5). BET and ICP-OES measurements
were carried out to measure the surface area of the photocatalyst and the content of the metal
nanoparticles on TiO2 before and after 4 cycles of ADH reactions, respectively. Thus, no leaching
was detected by ICP-OES measurements, showing the same amount of Rh nanoparticles (0.99
wt% with respect to TiO2) before and after four experimental runs. However, the surface area of
the catalyst was found to decrease from 316 m2/g to 272 m2/g after 4th cycle. Therefore, post
treatment of the catalyst was required to recover the specific surface area. For this purpose, the
water suspended photocatalyst was illuminated at ambient temperature for one hour under 500 W
solar simulator lamp in an oxygen atmosphere. As a result, the surface area was recovered up to
312 m2/g and the product yield also reverted to 96 % in 5th cycle. Moreover, the subsequent
experimental run showed similar yields to those obtained in the second run, which is clear evidence
of the good reusability of the catalyst. A similar trend was also observed regarding the
photocatalytic performance, even when the hard washing procedure with acetone is applied after
each successive run. Washing with acetone did not affect the surface area of the photocatalysts
which remained still reduced. However, this procedure resulted in the leaching of Rh metal
nanoparticles down to 0.72 wt% with respect to TiO2. Nevertheless, the reaction yield was still
almost the same as the results of the non-washing procedure and could be increased up to 96%
after light treatment (Figure S1). These results surely support the significant role of the surface
porosity for the catalytic performance as well as the importance of the surface recovery process.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
According to the achieved results, a reaction mechanism was proposed for the acceptorless
dehydrogenation (Scheme 5). From the UV-vis spectroscopy results, it became obvious that 1 wt%
Rh/TiO2 absorb light only in the UV region. Similarly, the employed amines have absorption in
the UV region. However, mixing the amine with 1 wt% Rh/TiO2 exhibited a significant redshift in
ACS Paragon Plus Environment
22

Page 23 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
the absorption spectrum up to ~500 nm, suggesting the formation of a visible-light-responsive
surface complex. Effective ADH reactions over Rh/TiO2 suggested that visible light activation did
not comply with the classic photo-excitation mechanism of the semiconductor with photon energy
equal to or greater than its bandgap energy. Instead, the photosensitization of TiO2 nanoparticles
with N-heterocycles is presumably occurred through the LMCT effect due to the Lewis acid-base
interaction of the semiconductor and the N-heterocycle. A similar phenomenon was reported by
Chen et al, where tertiary amine and TiO2 generate surface complexation resulting in redshift of
the absorption band of TiO2 at about 10 nm.54 As proposed in Scheme 5, upon visible light
excitation of the surface complex, the electron from the ground state of the amine is injected
directly into the conduction band of the TiO2 UV-100. Subsequently, noble metal co-catalysts
temporarily store injected electrons from the conduction band of TiO2. In the next step, the
photogenerated amine radical cation undergoes the oxidation reaction via two proton abstraction
to form the C=N bond. The subsequent reduction of two proton by electrons yield one hydrogen
molecule. The repeating process drives the formation of another hydrogen molecule and the
complete aromatization to yield quinoline.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSIONS:
In summary, commercially available TiO2 UV-100 based heterogeneous photocatalyst was
successfully applied for the acceptorless dehydrogenation of N-heterocycles for the first time. 1
wt% Rh0 nanoparticles photodeposited TiO2 and 2-propanol were found to be the best
photocatalyst and dispersant for these transformations, respectively. Dehydrogenation of broad
substrate scope of saturated N-heterocycles such as tetrahydroquinolines, tetrahydroisoquinolines,
indolines and 4,5-dihydroimidazolines were afforded in good to excellent yield in an argon
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 44
1
2
3
4
5
6
7
8
9
atmosphere. The robustness of the developed method was confirmed by sixfold subsequent
experimental runs. The present study shows that the perhydroquinoline substrate lacking aromatic
ring attached to the piperidine scaffold failed full aromatization, which proves the essential role of
the benzene ring in the dehydrogenation reaction. Remarkably, it was discovered that the
aromatization of 8-acetyl-5,6,7,8-THQ could be realized in the optimized reaction condition. This
finding opens a new pathway in the visible-light-mediated ADH of benzofused systems. Apart
from the synthesis of industrially significant organic compounds, the concomitant formation of
hydrogen molecules makes the developed method even more attractive from the viewpoint of
hydrogen storage application.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
FIGURES
ACS Paragon Plus Environment
24

Page 25 of 44
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Normalized extinction spectra of 1 wt% Rh/TiO2 (UV-100) suspension containing
1,2,3,4-THQ, 1,2,3,4-THisoQ, indoline in 2 mL i-C3H7OH (b) and F(R) spectra of noble metal
photodeposited TiO2 (UV-100) (a).
Figure 2. The influence of the photocatalysts amount in the time-dependent ADH of 1,2,3,4-THQ.
Reaction conditions: catalyst (1 wt% Rh/TiO2), 1,2,3,4-THQ (0.2 mmol), i-C3H7OH (2 mL), blue
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 44
1
2
3
4
LED light (λmax=453 nm), Argon atmosphere. The yields were determined by GC-FID using
5
6
benzyl alcohol (50 mol%) as the internal standard.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. The influence of the loading amount of Rh on TiO2 on ADH of 1,2,3,4-THQ. Reaction
conditions: Rh/TiO2 photocatalyst (5 mg), 1,2,3,4-THQ (0.2 mmol), i-C3H7OH (2 mL), blue LED
light (λmax=453 nm), Argon atmosphere, 18 h. The yields were determined by GC-FID using
benzyl alcohol (50 mol%) as the internal standard. The amount of producing hydrogen molecules
were determined by GC-TCD according to the calibration curve.
ACS Paragon Plus Environment
26

Page 27 of 44
ACS Catalysis
1
2
3
4
5
6
Figure 4. Normalized absorption and extinction spectra of 8-Acetyl-5,6,7,8-THQ and 1 wt%
Rh/TiO2 (UV-100) suspension containing 8-Acetyl-5,6,7,8-THQ in 2 mL i-C3H7OH, respectively.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Acceptorless dehydrogenation of 1,2,3,4-tetrahydroquinoline: recycling of 1 wt%
Rh/TiO2. Reaction conditions: catalyst (1 wt% Rh/UV-100), 1,2,3,4-THQ (0.2 mmol), i-C3H7OH
(2 mL), blue LED light (λmax=453 nm), Argon atmosphere. The yields were determined by GC-
FID using benzyl alcohol (50 mol%) as the internal standard. The amount of producing hydrogen
molecules were determined by GC-TCD according to the calibration curve. 5th and 6th
experimental runs were performed after recovering the surface of the photocatalyst.
SCHEMES
+ 2H2
N
N
H
H
H
N
N
+ H2
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 44
1
2
3
4
Scheme 1. Acceptorless dehydrogenation of 1,2,3,4-tetrahydroquinoline (top) and indoline
5
6
(bottom).
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Acceptorless decarboxylation and dehydrogenation mechanism of 2-carboxyl-indoline
upon visible-light illumination. Reaction conditions: 1 wt% Rh/TiO2 photocatalyst (5 mg), 2-
carboxyl-indoline (0.2 mmol) i-C3H7OH (2 mL), blue LED light (λmax=453 nm), Argon
atmosphere, 48 h.
ACS Paragon Plus Environment
28

Page 29 of 44
ACS Catalysis
1
2
3
4
5
6
N
N
1 wt% Rh/TiO2
N
+ H2
N
H
H
Blue LED 453 nm
48 h, i-PrOH
94 %,a 96 % H2
2t
1t
7
8
9
H
N
N
1 wt% Rh/TiO2
+ 2H2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Blue LED 453 nm
48 h, i-PrOH
N
N
H
1u
2u
N.D. ,3 % H2
H
N
N
+
N
N
1 wt% Rh/TiO2
+ nH2
n=1-5
+
Blue LED 453 nm
48 h, i-PrOH
2v*
0%
2v
4 %
1v
23 % H2
17%
O
O
1 wt% Rh/TiO2
N
N
Blue LED 453 nm
48 h, i-PrOH
+ 2H2
1w
2w
62 %,a 62 % H2
HO
OO
HO OO
O
O
1 wt% Rh/TiO2
N
N
+ H2
Blue LED 453 nm
48 h, i-PrOH
1x
2x
5 % yield , 6 % H2
Scheme 3. Acceptorless dehydrogenation of other N-heterocycles over 1 wt% Rh photodeposited
TiO2. Reaction conditions: 1 wt% Rh/UV-100 photocatalyst (5 mg), substrate (0.2 mmol) in i-
C3H7OH (2 mL), blue LED light (λmax=453 nm), Argon atmosphere, 48 h. Yield calculated
according to the NMR signals. The amount of evolved hydrogen molecules was determined by
GC-TCD according to the calibration curve. a Yield is that of the isolated product.
ACS Paragon Plus Environment
29