Communication
analysis), the above optimized conditions were applied
on the reaction mixture either under O2 or under air. To
our delight, both cases gave even better results than the
corresponding reactions directly from cinnamic acid
(Table 1, entries 6 and 8 vs Table 2, entries 1 and 2). How-
ever, when benzaldehyde (4a), malonic acid, piperidine,
CuOTf, FeCl3, Et3N and diethyl H-phosphonate were all
mixed together in DMSO and heated at 708C for 12 h, the
desired product 3a was only obtained in 60% GC yield.
Based on the results, we decided to employ the two-step,
one-pot procedure given in Table 2, entry 2 as our stan-
dard optimized conditions.
Table 1. Selected optimization of the reaction parameters between cinnamic
acid (1a) and H-phosphonate (2a).[a]
Entry
Fe salt
Cu salt
Et3N
Atmosphere
T [8C]
Yield [%][b]
[equiv]
1
2
3
4[c]
5[d]
6
7
8
9
10
11
12
FeCl3
FeBr3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
FeCl3
–
CuOTf
1
1
1
1
1
1
1
1
1
1
1
–
O2
O2
O2
O2
O2
O2
O2
Air
N2
O2
O2
O2
60
60
60
60
60
70
80
70
70
70
70
70
76
47
44
60
62
82(73)
65
82(75)[e]
NR
ND
ND
CuOTf
Cu(OTf)2
CuOTf
CuOTf
CuOTf
CuOTf
CuOTf
CuOTf
CuOTf
–
With the optimized conditions in hand, we investigated
the substrate scope (Scheme 2). Gratifyingly, this one-pot
domino strategy proved very effective for a variety of aro-
matic aldehydes and H-phosphonates, affording the de-
sired products in good yields (Scheme 2). In addition to
alkyl and alkoxy groups, such as methyl (3ba, 3ca, 3 fa),
ethyl (3da), tert-butyl (3ea), and methoxy (3ga), halo-
substituted benzaldehydes were also compatible with the
standard conditions, leading to halo-substituted b-keto-
phosphonates, which might be used for further structural
manipulation (3ha–3ka). As strong electron-withdrawing
groups, both trifluoromethyl and cyano-substituted aro-
FeCl3
FeCl3
CuOTf
ND
[a] Reaction conditions: Cinnamic acid (1a; 0.3 mmol), H-diethyl phosphonate
(2a) (4 equiv), Cu salt (5 mol%), and Fe salt (10 mol%) in DMSO (0.5m), Et3N
(1 equiv), atmosphere, 8 h. [b] GC yield. [c] With 2 equivalents of 2a. [d] With
3 equivalents of 2a. [e] Values in parentheses refer to yield of isolated product.
NR=no reaction; ND=not detected.
atmosphere at 608C, the desired product b-ketophosphonate
was formed in 76% yield (isolated product; Table 1, entry 1).
Further screening of a variety of copper salts including CuBr,
CuBr2, Cu(OTf)2, Cu(TFA)2 (TFA=trifluoroacetate), Cu2O, and
Cu(OAc)2 established that CuOTf is optimal (Table 1, entry 1
and Table S1, entries 1–6 in the Supporting Information). Other
Table 2. Screening of conditions for one-pot reaction between benzalde-
hyde (4a), malonic acid, and H-phosphonate (2a).
.
iron salts, such as FeBr3, FeSO4 7H2O, FeCl2, Fe(acac)2, Fe2(SO4)3,
FeBr2 exhibited no or low catalytic activity for the reaction.[14]
Screening of bases and solvents indicated that triethylamine
and DMSO were the best choices (Table 1, entry 1), whereas
other bases (KOH, (iPr)2NEt, DBU, TMG, Na2CO3 and tBuOK) and
solvents (DMF, DMP, MeCN, toluene, DCE and dioxane) were all
less effective for this transformation.[14] When the temperature
was increased to 708C, the desired product 3aa was obtained
in 82% GC yield and isolated in 73% yield (Table 1, entry 6). In-
terestingly, increasing of the reaction temperature to 808C did
not lead to a corresponding increase in the product yield; in-
stead, 3aa was obtained in only 65% yield (Table 1, entry 7).
When O2 was replaced by N2, no desired product 3aa was
formed (Table 1, entry 9), implying that O2 is also a vital prereq-
uisite for this reaction. Remarkably, the reaction proceeded just
as well in air atmosphere as it did in O2 atmosphere (Table 1,
entry 8). When the reactions were performed in the absence of
either iron salt (Table 1, entry 10), or copper salt (Table 1,
entry 11), or without base,[14] trace or very low yields of desired
product 3aa were obtained, demonstrating that FeCl3, CuOTf
and Et3N are all essential to this oxyphosphorylation and they
must work cooperatively to effect this transformation.
Entry
Reaction steps
Atmosphere
Yield [%][b]
1
2
3
2
2
1[d]
O2
air
air
84
85 (79)[c]
60
[a] Reaction conditions: a) Benzaldehyde (4a; 0.3 mmol), malonic acid
(0.5 mmol), piperidine (0.2 mmol), 1008C, DMSO (1 mL), overnight; b) H-di-
ethyl phosphonate (2a; 4 equiv), CuOTf (5 mol%), and FeCl3 (10 mol%) in
DMSO (0.5m), Et3N (1 equiv), 708C, atmosphere, 8 h. [b] GC yield.
[c] Values in parentheses refer to yield of isolated product. [d] Steps (a)
and (b) were mixed up together and the reaction heated at 708C for 12 h.
matic aldehydes performed well under the standard conditions
and the corresponding CF3- and CN-substituted b-ketophosph-
onates (3la and 3ma) were obtained to further increase the
versatility of this transformation with a view to accessing prod-
ucts to desired in the medicinal and synthetic chemistry com-
munities. Notably, this reaction is chemoselective (aldehyde C=
O vs. ketone C=O), since the acetyl group was tolerated in the
one-pot reaction conditions and gave the desired product in
reasonable yield (3na). Polycyclic aromatic aldehydes, such as
1-naphthylaldehyde (4o) and 2-naphthylaldehyde (4p), reacted
well and gave the corresponding b-ketophosphonates in 55%
and 61% yields respectively (3oa and 3pa). Heteroaromatic al-
dehydes, such as 2-thiophenecarboxaldehyde (4q), worked
Subsequently, we explored the domino one-pot reaction by
combining benzaldehyde (4a), malonic acid, and piperidine in
DMSO at 1008C.[15] Once the reaction reached completion (cin-
namic acid was formed and detected by TLC, GC, and GCMS
Chem. Eur. J. 2015, 21, 10654 – 10659
10655
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim