2
T. Hering, B. K o€ nig / Tetrahedron xxx (2016) 1e5
electrophilic chlorination, but rather on the reduction to induce
The N-chloro compounds with benzyl groups 1c and with ‘push-
pull’ substituents 1d (entries 3, 4) gave comparable yields,
but are unfavourable with respect to atom economy. Further opti-
mization of the reaction was therefore carried out using 1-
chloromorpholine (1b).
Having identified the suitable N-chloramine, we continued with
varying the solvent of the reaction (Table 2). All polar solvents and
DCM (entries 1e4) showed conversion to the desired product 4b,
whereas the non-polar solvents toluene and 1,2-dichloroethane
(1,2-DCE) gave only minor product formation (entries 5, 6). This
observation can be explained by the different solubility of the
photocatalyst, which poorly dissolves in non-polar solvents. The
reaction was most efficient in a mixture of MeCN and water (entry
4), which even led to double chlorination. The enhanced reactivity
when water is added to the reaction presumably results from
a higher solubility of peroxodisulfate ensuring a quicker re-
generation of the photocatalyst.
ꢀ
a cleavage of Cl yielding a nitrogen atom centered radical, which
reacts further to form a CeN-bond.17,18 Even though in some ex-
amples chloride is later incorporated into the product, it always
reacts as a nucleophile and not as an electrophile. To investigate
whether N-chloramines (1) can also be activated by photocatalytic
E
oxidation to undergo electrophilic aromatic substitution (S Ar), we
chose the chlorination of an electron rich arene, namely dime-
thoxybenzene (3a) as a model reaction (Scheme 2): A solution of 3a
(
[
0.25 mmol), the N-chloramine 1a (1.2 equiv), the photocatalyst
Ru(bpy) ]Cl (5 mol %) and ammonium peroxodisulfate (1.2 equiv)
3
2
to reoxidize the photocatalyst in MeCN was irradiated under N
2
-
atmosphere over night with blue LEDs (
tocatalytic reaction yielded 13% of the chlorinated arene 4a
l
max¼455 nm). The pho-
whereas without irradiation no chlorination was observed.
For further studies the substrate was changed from dimethox-
ybenzene (3a) to the less electron rich anisole (3b) since this will
circumvent the problem of double chlorination. Table 3 summa-
rizes the results of the reaction optimization. First, we continued
the solvent screening and tested different MeCN/H
2
O ratios (Table
Scheme 2. Test reaction for the photocatalytic chlorination of dimethoxybenzene (3a)
with 1-chlororpiperidine (1a).
3, entries 1e3). The reaction using a 4:1 mixture of MeCN and
2
H O showed an excellent yield of 95%. Using this solvent mixture
the catalyst loading could be lowered to 2 mol % without a change
in the yield. A further decrease to 1 mol % showed a slightly de-
creased yield of 80% (entries 2, 4, 5). Neither the addition of base
Even though the yield of chlorination was low, these initial re-
sults showed that N-chloramines can be activated for S Ar by
E
photocatalytic oxidation. Next, the reaction conditions were opti-
mized. First we investigated whether the metal based photocatalyst
Table 1
[
Ru(bpy)
3
]Cl
2
can be replaced by cheap organic dyes such as eosin Y
Variation of the N-chloro compound using the reaction conditions depicted in
þ
or 9-mesityl-10-methylacridinium perchlorate (Acr eMes)
Scheme 2
a
0
ꢁ
(
Scheme 3). The redox properties of eosin Y (E (EY*/EY ꢀ)¼0.79 V vs
Entry
N-chloro compound
Yield (%)b
13
Conversion (%)b
26
0 19
SCE) are similar to [Ru(bpy)
3
]Cl
2
(E (Ru(II)*/Ru(I))¼0.77 V vs SCE),
nevertheless only traces of chlorination could be obtained when
using 10 mol % eosin Y instead of [Ru(bpy)
3
]Cl
2
(yield 4a<5%). Next
1
2
3
þ
we tested Acr eMes, which is a very strong oxidant in its excited
state (E0(MAþ*/MA )¼2.08 V vs SCE), but despite this high oxidative
ꢁ
20
power the catalyst was less efficient in this transformation than
74
>99
>99
[
3 2
Ru(bpy) ]Cl (yield 4a 7%). A possible complication with this cat-
alyst could be that its oxidation potential is sufficiently high to
oxidize dimethoxybenzene (3a) directly and thus leads to un-
desired side reactions.21
62
4
61
>99
a
Reactions were carried out using 0.25 mmol 3a, 1.2 equiv of the respective N-
, and 5 mol % [Ru(bpy) ]Cl O in 1.5 mL
ꢂ6H
max¼455 nm) was 16 h.
Determined by GC analysis using anisole as the internal standard.
chloro compound, 1.2 equiv (NH
MeCN. The irradiation time (
4
)
2
S
2
O
8
3
2
2
l
b
Scheme 3. Employed photocatalyst and the wavelength used for irradiation.
Table 2
Results of the solvent screening
a
Entry
Solvent
Yield (%)b
Conversion (%)b
Next, we varied the N-chloramine to investigate the influence of
the substituents on the nitrogen. The results of this screening are
depicted in Table 1. The highest yields of chlorinated dimethox-
ybenzene 4a were obtained using 1-chloromorpholine (1b) (74%,
entry 2), which is except for the heteroatom structurally very similar
1
2
3
4
5
6
MeCN
MeOH
DCM
MeCN/H
Toluene
1,2-DCE
74
42
45
>99
60
45
>99
<5
22
c
2
O 3:1
48þDC
<5
7
y
to the previously employed piperidine derivative (1a, entry 1).
a
Reactions were carried out using 0.25 mmol 3a, 1.2 equiv 1b, 1.2 equiv
, and 5 mol % [Ru(bpy) ]Cl O in 1.5 mL of the respective solvent.
ꢂ6H
The irradiation time (
(
NH
4
)
2
S
2
O
8
3
2
2
l
max¼455 nm) was 16 h.
y
b
In reported aromatic chlorinations with N-chloramines the morpholine de-
Determined by GC analysis using anisole as the internal standard.
c
rivative showed higher efficiency than N-chloropiperidine, see Ref. 14.
DC¼double chlorination.