L.-Y. Jiao, et al.
MolecularCatalysis494(2020)111120
Table 1
Identification of the reaction conditions for the chitosan supported copper catalyst on the deaminated esterification of DPPAa.
Entry
Catalyst (x)
Additive
Solvent
T (oC)
t (h)
1
2
3
4
5
6
7
8
CS@CuSO4 (10)
CS@Cu(OTf)2 (10)
CS@CuCl (10)
CS@Cu (10)
CS@CuCl (10)
CS@CuCl (10)
CS@CuCl (5)
CS@CuCl (10)
CS@CuCl (10)
CS@CuCl (10)
bare CS
e
n-PrOH
n-PrOH
n-PrOH
n-PrOH
n-PrOH
n-PrOH
n-PrOH
n-PrOH
40
40
40
40
60
80
80
80
80
80
80
80
80
24
24
24
24
24
24
40
24
24
24
24
24
24
< 5
8
10
< 5
31
90
76
79
94
87
trace
5
9
e
e
e
trace
28
84
74
71
90
81
17
90
74
e
e
e
NaOH
Na2CO3
NaHCO3
Na2CO3
Na2CO3
Na2CO3
9
n-PrOH
n-PrOH
n-PrOH
10
11
12
13
25
92
77
CS@CuCl (10)
CS@CuCl (10)
a
Reaction conditions: DPPA (1a, 0.5 mmol, 1.0 equiv.), n-propyl alcohol (2a, excess amount), additive (0.25 mmol, 0.5 equiv.), catalyst (indicated amount),
tetracosane as internal standard, solvent (2.0 mL, 0.25 M) in a reaction tube under air atmosphere.
b
Determined by gas chromatography (GC) analysis.
Isolated yield through silica gel based flash column chromatography.
Volume ratio: 9:1.
c
d
form a new PeO bond as expected, although the desired product 3aa,
namely diphenyl propyl phosphate, was generated in very low con-
versions and yields at 40 °C (entries 1–4). It was shown that CS@CuCl
gave the best performance among these candidates (entry 3). The mo-
lecular structure was confirmed unambiguously through NMR (nuclear
magnetic resonance) and HRMS (high resolution mass spectroscopic)
analyses. Then several parameters such as reaction temperatures, ad-
ditives, and reaction mediums were investigated. It has been proved
that the reaction temperature played an important role in the reactivity.
Elevating temperature from 40 °C to 80 °C could provide essential im-
provement of the isolated yields (entries 3, 5, and 6). Notably, the
deaminated esterification underwent smoothly even with catalyst
loading reduced by half, albeit with a lower yield and much longer
reaction time (entry 7). Additives such as NaOH, Na2CO3, and NaHCO3
were evaluated to alter catalytic system, in which Na2CO3 obviously
exhibited positive effect thereof (entries 8–10). Interestingly, the bare
chitosan itself also showed low catalytic activity towards the model
reaction (entry 11), providing promising evidence in this topic. Con-
cerning the generality of our protocol and the increased solvent visc-
osity of other alcohols, we decided to introduce other solvent to form
mixed reaction media. Non-polar solvent for example cyclohexane
proved to be an ideal choice, giving rise to 90 % yield (entry 12).
Therefore, the optimal conditions was performed with the aid of 10 mol
% of CS@CuCl as catalyst, 0.5 equivalent of Na2CO3 as additive at
80 °C, under which either aliphatic alcohol alone or cyclohexane con-
taining reaction medium performed equally well (entries 9 and 12).
Importantly however, it was noteworthy that the isolated yields and
conversions correlated very well, we did not detect any other by-pro-
duct in this system with almost unreacted DPPA recovery (isolated after
purification by flash column chromatography on silica gel, not shown).
With the optimized catalytic conditions established, we were next
promoted to investigate various phosphoryl azides 1a–1k and aliphatic
alcohols 2a–2s with different substituents (Table 2). Not to our sur-
prise, moderate to excellent results were obtained for all the straight
chain fatty alcohols from C1 to C8 (2a–2c → 3aa–3ac, 2g → 3ag, 2i →
3ai, 2k–2m → 3ak–3am) (entries 1–3, 7, 9, and 11–13). Gratifyingly,
methanol (2b) and ethanol (2c) represented highly reactive nucleo-
philic reagents, providing full conversions and 90–92 % yields within
3 h (entries 2–3). Moreover, the fluorinated derivatives trifluoroethanol
(TFE, 2d) and hexafluoroisopropanol (HFIP, 2f) afforded corresponding
products in 84 % and 96 % yield, respectively (entries 3–6). When
cycloalkyls substituted alcohols such as cyclopentanol (2n) and cyclo-
hexanol (2o) were investigated, decreased results were obtained (en-
tries 14 and 15). It might be attributed to the steric hindrance effect on
the SN2 attack process. This hypothesis was convinced immediately
with less bulky substrates 2p and 2q containing at least one methylene
(CH2) group, and good results were realized though (entries 16 and 17).
Remarkably, this protocol was not applicable for either benzyl alcohol
(2r) or phenethyl alcohol (2s) bearing similar functional groups,
however, only traces amounts of products were observed in these aro-
matic substrates (entries 18 and 19).
Subsequently, we then turned our attention to examine different
phosphoryl azides under identical reaction conditions with methanol
(2b) as the oxygen source (Table 2). These reagents have been de-
monstrated to be effective under the heterogeneous catalysis, providing
methyl diaryl phosphate derivatives in moderate to excellent yields.
Generally, functional groups commonly existing in organic synthesis
such as methyl (1b–1d), other alkyl (1g and 1h), methoxyl (1i and 1j),
were well tolerated. However, bromo (1e and 1f) substituted substrates
led to reduced yields (entries 20–28). Compared with phosphoryl azides
bearing single substituent on the aromatic core, disubstituted starting
material at both 3- and 5-positions had performed equally well, re-
gardless of the electronic or steric properties (1k → 3ka) (entry 29).
Furthermore, this heterogeneous catalysis was proceeded success-
fully in gram scale reaction with 86 % yield (5.0 mmol of 1a was em-
ployed as starting material), further demonstrated the practicality and
scalability of our protocol (Table 2, entry 1).
Based on the aforementioned results, we then extended the alcohols
to more complicate glycerol derivative 2t and citronellol (2u) (Scheme
2). Compared with desired product 3au in 84 % yield, however, alcohol
2t with a dimethylketal group seemed unstable under basic conditions,
and only moderate result was achieved finally.
3