Photocatalytic decarboxylative alkylations mediated by triphenylphosphine and sodium iodide

Article abstract of DOI:10.1126/science.aav3200

Most photoredox catalysts in current use are precious metal complexes or synthetically elaborate organic dyes, the cost of which can impede their application for large-scale industrial processes. We found that a combination of triphenylphosphine and sodium iodide under 456-nanometer irradiation by blue light–emitting diodes can catalyze the alkylation of silyl enol ethers by decarboxylative coupling with redox-active esters in the absence of transition metals. Deaminative alkylation using Katritzky’s N-alkylpyridinium salts and trifluoromethylation using Togni’s reagent are also demonstrated. Moreover, the phosphine/iodide-based photoredox system catalyzes Minisci-type alkylation of N-heterocycles and can operate in tandem with chiral phosphoric acids to achieve high enantioselectivity in this reaction.

Full text of DOI:10.1126/science.aav3200

RESEARCH
sive dyes. Specifically, we targeted light-induced
intermolecular electron transfer from sodium
iodide to an aliphatic redox-active ester (RAE)
(24–26) to induce radical decarboxylation con-
trollably and thereby deliver an alkyl radical useful
in organic synthesis (Fig. 1B).
ORGANIC CHEMISTRY
Photocatalytic decarboxylative
alkylations mediated by
Simulations of charge
transfer energetics
triphenylphosphine and sodium iodide
Sodium iodide is known to reduce aryl bromide
and triflate to the respective aryl radicals for the
aromatic Finkelstein reaction, but only under
high-energy ultraviolet (UV) irradiation (27, 28).
Through density functional theory (DFT) calcula-
tions (see supplementary materials), we estimated
that electron transfer from sodium iodide to
the RAE N-(cyclohexanecarbonyl)phthalimide is
endergonic by 56.2 kcal/mol, but only by 44.3 kcal/
mol in the presence of triphenylphosphine (PPh3)
because of the favorable formation of the Ph3P–I•
radical (calculated to be exergonic by 11.9 kcal/
mol) (29, 30). A Ph3P–I• species was observed by
electron paramagnetic resonance spectroscopy
in the 1970s (30). Theoretical calculations (see
supplementary materials) and natural bond
orbital analysis suggest a reduction potential of
0.69 V versus saturated calomel electrode (SCE),
and spin densities delocalized across I and P are
0.44 and 0.42, respectively (Fig. 1B). The calcu-
lations also suggested that complexation of NaI
and PPh3 is exergonic in acetonitrile through the
cation-p interaction (exergonic by 4.6 kcal/mol).
The assembly of NaI and PPh3 with the RAE
N-(cyclohexanecarbonyl)phthalimide to form a
charge transfer complex (CTC) via coulombic inter-
action is calculated to be exergonic by 3.8 kcal/mol
Ming-Chen Fu1, Rui Shang1,2*, Bin Zhao1, Bing Wang1, Yao Fu1*
Most photoredox catalysts in current use are precious metal complexes or synthetically
elaborate organic dyes, the cost of which can impede their application for large-scale
industrial processes. We found that a combination of triphenylphosphine and sodium
iodide under 456-nanometer irradiation by blue light–emitting diodes can catalyze the
alkylation of silyl enol ethers by decarboxylative coupling with redox-active esters in the
absence of transition metals. Deaminative alkylation using Katritzky’s N-alkylpyridinium
salts and trifluoromethylation using Togni’s reagent are also demonstrated. Moreover,
the phosphine/iodide-based photoredox system catalyzes Minisci-type alkylation of
N-heterocycles and can operate in tandem with chiral phosphoric acids to achieve
high enantioselectivity in this reaction.
he power of light-induced electron trans-
fer for catalytic organic synthesis (1–8)
has been demonstrated by the remarkable
recent progress in photoredox catalysis (9, 10).
The photoredox catalysts in current use are
in photochemistry (Fig. 1A), which may not re-
quire each substrate (donor or acceptor) to absorb
at the desired wavelength individually (20, 21).
Such irradiation-induced intermolecular charge
transfer is applied in organic photovoltaics (22)
and traditional photochemistry (23) but is seldom
used as a principle to construct a catalytic redox
cycle for organic synthesis.
We posited that a photoredox catalytic cycle
based on light-induced intermolecular electron
transfer without direct excitation of the catalyst
and substrates could obviate the need for expen-
T
mostly precious metal complexes (11–15) and syn-
thetically elaborate organic dyes (16), which have
charge-separated excited states (17–19) accessible
by absorption of visible light (Fig. 1A). Photo-
induced intermolecular charge transfer through
assembly of donor and acceptor molecules by
noncovalent interactions is a well-known process
Fig. 1. Redox catalysis based on light-induced
intermolecular electron transfer from sodium
iodide to redox-active ester. (A) Light-
induced intramolecular charge transfer
(CT) and light-induced intermolecular donor
(D)–acceptor (A) charge transfer through
self-assembly via noncovalent interactions.
L, ligand; M, metal. (B) Estimated Gibbs
energy change of intermolecular electron
transfer from NaI and NaI/PPh3 component
to aliphatic N-(acyloxy)phthalimide (NPhth) to deliver an alkyl radical, and a possible iodide/phosphine redox cycle. (C) Photoactivation of assembled
complex of N-(acyloxy)phthalimide with NaI and PPh3 through coulombic and cation-p interactions. SET, single electron transfer.
Fu et al., Science 363, 1429–1434 (2019)
29 March 2019
1 of 6

RESEARCH
| RESEARCH ARTICLE
(Fig. 1C). The energy barrier of the electron trans-
fer process from iodide to the phthalimide moiety
was estimated to be 61.2 kcal/mol using Marcus
theory, within 1.5 kcal/mol of the photon en-
ergy that 456-nm blue LEDs can provide. The
analogous electron transfer process in the ab-
sence of PPh3 must overcome a higher barrier
of 86.5 kcal/mol (Fig. 1C; see supplementary
materials). Further computational studies on
the excited state of the CTC assigned the S0-to-
S1 excitation to electron transfer from iodide to
the p* orbital of the phthalimide moiety with an
excitation energy of 2.85 eV, which corresponds
to a wavelength of 436 nm (fig. S3). On the basis
of the above theoretical analysis, we explored a
simple combination of NaI and PPh3 as a photo-
redox catalyst for decarboxylative alkylation re-
actions (31, 32).
gave comparable yields. However, highly electron-
deficient tris(4-pentafluorophenyl)phosphine
was completely ineffective, probably due to its
lack of electron-donating capacity to facilitate
electron transfer of the iodide salt. The use of
tricyclohexylphosphine lowered the yield, as
did the sterically bulky triarylphosphine ligand
2-(diphenylphosphino)biphenyl, which likely
hindered formation of the CTC (Fig. 2A). All
three components—phosphine, sodium iodide,
Investigation of key reaction parameters
The optimized reaction conditions for decarbox-
ylative alkylation using NaI/PPh3 are shown in
Fig. 2. Decarboxylative cyclohexyl addition to
trimethyl[(1-phenylvinyl)oxy]silane delivered
a-cyclohexylacetophenone in 82% yield under blue
LED irradiation of 20 mole percent (mol %) PPh3
and 150 mol % NaI in acetonitrile (see table S1
for details of optimization) (33, 34). The reaction
requires further desilylation by a base to form
the a-alkylated ketone. Because NaI is so in-
expensive, it was used in superstoichiometric quan-
tity (1.5 equivalent) both as electron-transfer catalyst
and base to trap the trimethylsilyl (TMS) cation.
Extremely pure (99.999%) NaI without any
metal contamination was tested and gave the
same results observed with the commonly avail-
able reagent-grade material (purity >99.0%). The
results of testing other alkali halides are shown in
Fig. 2A. Lithium and potassium iodide were much
less effective than NaI, and a soluble quaternary
ammonium iodide was entirely ineffective. The
observed alkali metal cation effect revealed that
the sodium cation has an important role in the
electron transfer activation step, as indicated by
DFT study in Fig. 1C (formation of the CTC by LiI,
NaI, and KI is exergonic by 1.1 kcal/mol, 3.8 kcal/
mol, and 2.9 kcal/mol, respectively, as indicated
by DFT calculation). Other sodium halides (fluo-
ride, chloride, and bromide) were also ineffective.
As noted above, phosphine is crucial to facilitate
intermolecular charge transfer and stabilizes the
iodine radical as a R3P–I• species (30). Thus, we
also screened a series of phosphines with different
electronic and steric properties (Fig. 2A, second
row). The results showed that the electronic proper-
ties of triarylphosphines did not significantly affect
the reaction efficiency, as using tris(4-fluorophenyl)
phosphine and tris(4-methoxyphenyl)phosphine
1Hefei National Laboratory for Physical Sciences at the
Microscale, CAS Key Laboratory of Urban Pollutant
Conversion, Anhui Province Key Laboratory of Biomass Clean
Energy, iChEM, University of Science and Technology of
China, Hefei, Anhui 230026, China. 2Department of
Chemistry, School of Science, University of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan.
*Corresponding author. Email: rui@chem.s.u-tokyo.ac.jp (R.S.);
fuyao@ustc.edu.cn (Y.F.)
Fig. 2. Key reaction-controlling parameters of NaI and PPh3–catalyzed decarboxylative
alkylation. (A) Parameters affecting decarboxylative alkylation of silyl enol ethers. (B) Parameters
affecting decarboxylative alkylation of N-heteroarenes. (C and D) UV-Vis absorption spectra
of reactant mixtures. Concentration of each substance in UV-Vis measurement is identical to
the concentration used in reactions. Me, methyl; TFA, trifluoroacetic acid.
Fu et al., Science 363, 1429–1434 (2019)
29 March 2019
2 of 6

RESEARCH
| RESEARCH ARTICLE
and irradiation—were essential for the reaction
(Fig. 2A). Moreover, addition of 10 mol % of iodine
to the reaction mixture entirely suppressed the
reaction (35). This poisoning effect rules out a
productive role for iodine or I3– in the redox cycle
and highlights the important role of PPh3 in
keeping the iodine radical trapped as the putative
persistent R3P–I• species. The classic CTC I2•PPh3
was also ineffective.
emitting at 520 nm were ineffective, whereas
blue and purple LEDs at wavelengths of 456 nm,
440 nm, and 427 nm were comparably effective.
Shorter wavelengths than these resulted in lower
yields. Investigation of reaction parameters for
decarboxylative Minisci-type alkylation (36, 37)
of 4-methylquinoline (Fig. 2B) gave results sim-
ilar to those observed for the alkylation of silyl
enol ether (Fig. 2A), except that for the Minisci
reaction, both sodium iodide and PPh3 could be
used in catalytic quantities (5 to 20 mol %; see
table S4 for a detailed parameter study). Sever-
al commonly used low-boiling solvents, such as
trifluoromethyltoluene, acetone, acetonitrile, and
dioxane, effectively dissolved 10 mol % of NaI and
afforded alkylation products in high yield (>85%).
Control experiments showed that for the Minisci
alkylation, PPh3, NaI, irradiation, and a Brønsted
acid additive to increase the electrophilicity of the
N-heteroarene were all essential (Fig. 2B). Once
again, addition of 10 mol % of iodine completely
suppressed the reaction. The quantum yield of
the decarboxylative Minisci alkylation (Fig. 2B)
was measured to be 0.15, a value suggesting a
closed catalytic cycle rather than a radical chain
process initiated by electron transfer (38, 39).
Cyclohexyl iodide was not detected in the re-
action mixtures of the silyl enol ether or the
N-heteroarene, and a control experiment using
cyclohexyl iodide as alkylation reagent did not
yield any desired alkylation products.
We next explored the dependence of the re-
action on irradiation wavelength. Across the
spectrum (Fig. 2A and table S1), green LEDs
To provide further understanding of the NaI/
PPh3 photoredox system, we measured UV-visible
(UV-Vis) absorption spectra. Reactant concen-
tration effects on the UV-Vis absorption spectrum
in a light-induced intermolecular donor-acceptor
charge-transfer reaction were reported by Miyake
and co-workers for C–S coupling of aryl halides
(40). We performed a UV-Vis absorption measure-
ment using a solution of the same concentration
as the real reaction mixture (Fig. 2, C and D), from
which it was apparent that neither NaI, PPh3,
nor the combination of NaI and PPh3 has any
absorption in the visible region (Fig. 2D). Silyl
enol ether and 4-methylquinoline showed ab-
sorption features only in the UV (<350 nm). The
redox-active ester showed an absorption onset
around 390 nm. The mixture of redox-active
ester (1) with either silyl enol ether (2) or 4-
methylquinoline (4) did not show any significant
change of the absorption onset compared with
redox-active ester (1) alone, which suggests that
no intermolecular charge transfer took place.
However, an obvious redshift of absorption
onset, tailing into the wavelength range of blue
LED irradiation, was observed when the NaI/
PPh3 component was mixed with redox-active
ester. This redshift supports the formation of a
charge-transfer complex (41, 42) between NaI/
PPh3 and redox-active ester in the reaction mix-
ture (see Fig. 1C).
Application to alkylation of silyl enol
ethers and N-heteroarenes
Next, we explored the reaction scope for ketone
synthesis through decarboxylative alkylation of silyl
enol ethers (Fig. 3). A broad range of functional
groups such as ether (6), alkyl chloride (7), ter-
minal alkyne (8), terminal alkene (9), ester (10),
amide (12), sulfide (23), aryl fluoride (27), aryl
bromide (28), aryl chloride (29), aryl iodide
(30), trifluoromethyl (33), acidic methyl sulfone
(34), and even aryl pinacol boronate (43) proved
compatible. Redox-active esters derived from var-
ious natural and unnatural amino acids were re-
active to give a-aminoalkylation products in good
yields. The scalability of this reaction was demon-
strated by preparation of 2.9 g of amino acid de-
rivative 18. The low cost of this NaI/PPh3 catalytic
Fig. 3. Scope for decarboxylative alkylation of silyl enol ether. Reaction conditions: redox-active
ester (0.2 mmol, 1.0 equiv), silyl enol ethers (0.4 mmol), NaI (0.3 mmol), PPh3 (20 mol %),
CH3CN (2 ml), 15 hours, room temperature (r.t.), 456-nm blue LEDs. Isolated yields are reported.
*Reaction on 8.0 mmol scale. Boc, tert-butyloxycarbonyl.
Fu et al., Science 363, 1429–1434 (2019)
29 March 2019
3 of 6

RESEARCH
| RESEARCH ARTICLE
system is appealing for industrial application to
large-scale syntheses.
The scope of Minisci-type alkylation (Fig. 4A)
spanned redox-active esters derived from sec-
ondary (46, 49) and tertiary (50–53) aliphatic
carboxylic acids, a-amino acids (54–59), a-hydroxy
acids (47, 48), and peptides (63) (36). The catalyst
loading for a gram-scale reaction could be re-
duced to 5 mol % of NaI and 5 mol % of PPh3 to
give 2.78 g of alkylation product 62 in 80%
yield. Besides 4-methyl quinoline, other sub-
stituted quinolines were also reactive (64, 70).
Alkylation took place on the C4 position of the
quinoline ring when C2-substituted quinolines
were tested (65, 71). Isoquinolines (66–68),
Fig. 4. Minisci-type decarboxylative alkylation. (A) Scope for Minisci-
type decarboxylative alkylation of N-heteroarenes. Reaction conditions:
N-heteroarenes (0.2 mmol, 1.0 equiv), redox-active ester (0.3 mmol),
NaI (10 mol %), PPh3 (20 mol %), TFA (0.2 mmol), acetone (2 ml),
15 hours, r.t., 456-nm blue LEDs. Isolated yields are reported.
*PhCF3 as solvent. †N-heteroarenes (8.0 mmol, 1.0 equiv),
redox-active ester (8.8 mmol, 1.1 equiv), NaI (5 mol %), PPh3
(5 mol %), TFA (8.0 mmol), acetone (40 ml), 15 hours, r.t., 456-nm
blue LEDs. ‡( )-1,1′-binaphthyl-2,2′-diyl hydrogenphosphate
(5.0 mol %) instead of TFA (0.2 mmol). (B) Merging NaI/PPh3
photoredox catalysis with chiral phosphoric acid (PA) catalysis
for enantioselective Minisci-type a-aminoalkylation. Reaction
conditions: N-heteroarenes (0.1 mmol, 1.0 equiv), redox-active
ester (0.15 mmol), NaI (20 mol %), PPh3 (20 mol %), chiral PA
(5.0 mol %), 1,4-dioxane (2 ml), 20 hours, r.t., 456-nm blue LEDs.
Isolated yields are reported; enantiomeric excesses were
determined by high-performance liquid chromatography (HPLC).
(C) Scope of enantioselective Minisci-type decarboxylative
a-aminoalkylation by relay of NaI/PPh3 redox catalysis with chiral
anion catalysis. Reaction conditions: N-heteroarenes (0.2 mmol,
1.0 equiv), redox-active ester (0.3 mmol), NaI (10 mol %),
PPh3 (10 mol %), (R)-TRIP-PA (5 mol %), 1,4-dioxane (2 ml),
15 hours, r.t., 456-nm blue LEDs. Isolated yields are reported;
enantiomeric excesses were determined by HPLC. Absolute
stereochemistry of products was assigned by analogy to 73.
*N-heteroarenes (0.3 mmol), redox-active ester (0.2 mmol). †NaI
(20 mol %), PPh3 (20 mol %), (R)-TRIP-PA (10 mol %). Cy, cyclohexyl;
Et, ethyl; t-Bu, tert-butyl; Ac, acetyl; i-Pr, isopropyl.
Fu et al., Science 363, 1429–1434 (2019)
29 March 2019
4 of 6

RESEARCH
| RESEARCH ARTICLE
Fig. 5. General applicability of NaI/PAr3
photoredox system. (A) Decarboxylative
alkenylation of RAEs. (B) Deaminative
alkenylation of Katritzky’s N-alkylpyridinium
salts. (C) Trifluoromethylation using
Togni’s reagent. *Using PPh3 instead of
P(p-MeO-C6H4)3. †Using P(p-F-C6H4)3
instead of P(p-MeO-C6H4)3. ‡Yield
1
measured by H nuclear magnetic
resonance using diphenylmethane as
internal standard.
Fig. 6. A proposed full catalytic
cycle of NaI/PPh3 photoredox catalysis.
Catalytic cycle of decarboxylative
alkylation with N-heteroarene is
demonstrated as an example.
See fig. S4 for proposed full catalytic
cycles for reactions with silyl enol
ether and alkene.
Fu et al., Science 363, 1429–1434 (2019)
29 March 2019
5 of 6

RESEARCH
| RESEARCH ARTICLE
phenanthridine (69), and pyridine (72) were
all effective substrates, yielding a variety of alkyl-
ated N-heteroarenes of pharmaceutical impor-
tance (43).
control experiments confirmed the essential roles
of NaI, PAr3, and irradiation. Solvent plays a
crucial role for these transformations [e.g.,
dimethylformamide (DMF) as solvent is crucial
for deaminative alkylation, as the reaction failed
in acetonitrile and acetone], probably because
noncovalent interactions required for assem-
bling the CTC, such as cation-p and electrostatic
interactions, are heavily influenced by solva-
tion. Last, a proposed full catalytic cycle of NaI/
PAr3 photoredox catalysis is illustrated in Fig. 6
by taking Minisci alkylation as an example (see
fig. S4 for proposed full catalytic cycles for re-
actions with silyl enol ether and alkene). After
photofragmentation of the CTC, the generated
alkyl radical attacks N-heteroarene to form a
carbon-carbon bond. The PPh3-I• radical oxi-
dizes the delocalized carbon radical generated
after the alkyl radical attacks the p system to
regenerate PPh3 and NaI. Generally, the oxida-
tion potentials of delocalized carbon radicals
(such as benzylic radical and allylic radical) are
lower than the reduction potential of PPh3-I•
(0.69 V versus SCE) (50). Thus, the redox po-
tential of PPh3-I• is sufficient to close the redox
cycle.
22. K. Leo, Nat. Rev. Mater. 1, 16056 (2016).
23. M. Oelgemöller, Chem. Rev. 116, 9664–9682 (2016).
24. K. Okada, K. Okamoto, N. Morita, K. Okubo, M. Oda, J. Am.
Chem. Soc. 113, 9401–9402 (1991).
25. T. Qin et al., Science 352, 801–805 (2016).
26. J. T. Edwards et al., Nature 545, 213–218 (2017).
27. L. Li et al., J. Am. Chem. Soc. 137, 8328–8331
(2015).
28. W. Liu, X. Yang, Y. Gao, C.-J. Li, J. Am. Chem. Soc. 139,
8621–8627 (2017).
29. Z. Guo et al., Org. Lett. 20, 1684–1687 (2018).
30. M. C. R. Symons, R. L. Petersen, J. Chem. Soc. Faraday Trans.
II 75, 210–219 (1979).
31. S. Bloom et al., Nat. Chem. 10, 205–211 (2018).
32. A. C. Sun, E. J. McClain, J. W. Beatty, C. R. J. Stephenson,
Org. Lett. 20, 3487–3490 (2018).
33. I. Kuwajima, E. Nakamura, Acc. Chem. Res. 18, 181–187
(1985).
34. W. Kong, C. Yu, H. An, Q. Song, Org. Lett. 20, 349–352
(2018).
35. F. A. Cotton, P. A. Kibala, J. Am. Chem. Soc. 109, 3308–3312
(1987).
36. W.-M. Cheng, R. Shang, Y. Fu, ACS Catal. 7, 907–911
(2017).
37. W.-M. Cheng, R. Shang, M.-C. Fu, Y. Fu, Chem. Eur. J. 23,
2537–2541 (2017).
38. A. Studer, D. P. Curran, Nat. Chem. 6, 765–773
(2014).
39. B. Zhang, C. Mück-Lichtenfeld, C. G. Daniliuc, A. Studer, Angew.
Chem. Int. Ed. 52, 10792–10795 (2013).
40. B. Liu, C. H. Lim, G. M. Miyake, J. Am. Chem. Soc. 139,
13616–13619 (2017).
41. Y. Cheng, C. Mück-Lichtenfeld, A. Studer, J. Am. Chem. Soc.
140, 6221–6225 (2018).
42. L. Candish, M. Teders, F. Glorius, J. Am. Chem. Soc. 139,
7440–7443 (2017).
43. E. Vitaku, D. T. Smith, J. T. Njardarson, J. Med. Chem. 57,
10257–10274 (2014).
44. R. J. Phipps, G. L. Hamilton, F. D. Toste, Nat. Chem. 4,
603–614 (2012).
Merging with chiral Brønsted acid
catalysis for enantioselective alkylation
To our excitement, we found that the NaI/PPh3
redox catalyst could operate synergistically with
a chiral Brønsted acid catalyst (44, 45) to achieve
asymmetric a-aminoalkylation of N-heteroarenes
(Fig. 4B). This enantioselective transformation was
reported only recently by Phipps and co-workers
(46) using an expensive iridium photoredox cata-
lyst, following from the precedent reported by
our group pairing the iridium with achiral phos-
phoric acid catalysis (36). Here, we combined
20 mol % of NaI/PPh3 with 5 mol % of chiral
phosphoric acid in the absence of transition
metals. Evaluation of various commercially avail-
able chiral phosphoric acids showed that (R)-
TRIP-PA [(R)-3,3′-bis(2,4,6-triisopropylphenyl)-1,
1′-binaphthyl-2,2′-diyl hydrogenphosphate] was
the optimal choice to deliver (S)-a-aminoalkylated
product in 97% yield and 95% enantiomeric ex-
cess (ee). Zhou-type spiro-phosphoric acids (47)
were also found to be effective, giving compa-
rable yield and enantioselectivity. The absolute
configuration of the a-aminoalkylated products
was unambiguously determined by x-ray single-
crystal analysis (73). The configuration is switch-
able by changing the absolute configuration of
the chiral phosphoric acid catalyst. A broad scope
of natural and unnatural a-amino acid–derived
RAEs was applicable to the asymmetric decar-
boxylative Minisci-type a-aminoalkylation re-
action (73–80) to produce various valuable
enantioenriched basic heterocycles in high enan-
tioselectivity (Fig. 4C). For quinoline deriva-
tives that did not possess substituents on the
2- or 4-positions, enantioselective alkylation
proceeded with C2 selectivity. Besides quinoline,
asymmetric decarboxylative C2-alkylations of
functionalized pyridines (84–86) were also
achieved in high yields and high enantiose-
lectivity. Isoquinoline was reactive to give the
a-aminoalkylation product in high yield, but the
enantioselectivity was only 33% ee (see supple-
mentary materials).
Because the noncovalent interaction (cation-p
interaction, Coulombic interaction, etc.) required
for assembly of CTC is rather common, and electron
transfer from iodide to many organic molecules is
precedented under UV (27, 28) or high-temperature
conditions (39), we posited that the iodide phos-
phine photoredox system should be generally ap-
plicable to substrates other than RAEs. Indeed,
besides decarboxylative alkenylation using RAEs
with 1,1-diphenylethylene (Fig. 5A), we have
found that NaI/PAr3 also activates Katritzky’s N-
alkylpyridinium salts to enable catalytic deam-
inative alkylation (48) with 1,1-diarylethylene to
deliver alkyl Heck-type products (49) (88–91)
(Fig. 5B). The NaI/PPh3 system also activated
Togni’s reagent for photoredox trifluoromethy-
lation of 1,1-diarylethylene and silyl enol ether
(92 and 93) (Fig. 5C). For all these reactions,
We hope the reactions presented above will
inspire future research in photoredox catalysis
by introducing a tricomponent catalytic system
based on a salt, a phosphine, and an electron-
accepting substrate to access the CTC without
the need of a traditional dye- or metal complex–
based photoredox catalyst.
45. D. Uraguchi, M. Terada, J. Am. Chem. Soc. 126, 5356–5357
(2004).
46. R. S. J. Proctor, H. J. Davis, R. J. Phipps, Science 360, 419–422
(2018).
47. J.-X. Guo, T. Zhou, B. Xu, S.-F. Zhu, Q.-L. Zhou, Chem. Sci. 7,
1104–1108 (2016).
48. J. Wu, L. He, A. Noble, V. K. Aggarwal, J. Am. Chem. Soc. 140,
10700–10704 (2018).
49. G.-Z. Wang, R. Shang, W.-M. Cheng, Y. Fu, J. Am. Chem. Soc.
139, 18307–18312 (2017).
50. Y. Fu, L. Liu, H.-Z. Yu, Y.-M. Wang, Q.-X. Guo, J. Am. Chem.
Soc. 127, 7227–7234 (2005).
REFERENCES AND NOTES
1. D. A. Nicewicz, D. W. C. MacMillan, Science 322, 77–80
(2008).
2. J. Twilton et al., Nat. Rev. Chem. 1, 0052 (2017).
3. J. Jin, D. W. C. MacMillan, Nature 525, 87–90 (2015).
4. M. H. Shaw, V. W. Shurtleff, J. A. Terrett, J. D. Cuthbertson,
D. W. C. MacMillan, Science 352, 1304–1308 (2016).
5. E. B. Corcoran et al., Science 353, 279–283 (2016).
6. I. Ghosh, T. Ghosh, J. I. Bardagi, B. König, Science 346,
725–728 (2014).
7. A. Bauer, F. Westkämper, S. Grimme, T. Bach, Nature 436,
1139–1140 (2005).
8. M. Silvi, C. Verrier, Y. P. Rey, L. Buzzetti, P. Melchiorre,
Nat. Chem. 9, 868–873 (2017).
9. M. H. Shaw, J. Twilton, D. W. C. MacMillan, J. Org. Chem. 81,
6898–6926 (2016).
10. C. R. J. Stephenson, T. P. Yoon, D. W. C. MacMillan, Eds.,
Visible Light Photocatalysis in Organic Chemistry (Wiley,
2018).
ACKNOWLEDGMENTS
Funding: Supported by National Key R&D Program of China
(2018YFB1501600, 2017YFA0303502), National Natural
Science Foundation of China (21572212, 21732006, 51821006),
Strategic Priority Research Program of CAS (XDB20000000,
XDA21060101), HCPST (2017FXZY001), KY (2060000119), and
the Supercomputing Center of USTC. Author contributions:
R.S. conceived the idea, guided the project, and wrote the
manuscript; M.-C.F. and B.Z. performed the experiments;
B.W. performed the theoretical study; and R.S., M.-C.F., and
Y.F. analyzed the data and participated in the preparation of
the manuscript. Competing interests: The authors declare
no competing financial interests. Data and materials
availability: Crystallographic data are available free of
charge from the Cambridge Crystallographic Database Centre
(CCDC 1891670). All other data are available in the main text
or the supplementary materials.
11. C. K. Prier, D. A. Rankic, D. W. C. MacMillan, Chem. Rev. 113,
5322–5363 (2013).
12. C. P. Johnston, R. T. Smith, S. Allmendinger, D. W. C. MacMillan,
Nature 536, 322–325 (2016).
13. E. R. Welin, C. Le, D. M. Arias-Rotondo, J. K. McCusker,
D. W. C. MacMillan, Science 355, 380–385 (2017).
14. C. Le, Y. Liang, R. W. Evans, X. Li, D. W. C. MacMillan, Nature
547, 79–83 (2017).
15. J. C. Tellis, D. N. Primer, G. A. Molander, Science 345, 433–436
(2014).
16. N. A. Romero, D. A. Nicewicz, Chem. Rev. 116, 10075–10166
(2016).
17. A. Vogler, H. Kunkely, Coord. Chem. Rev. 208, 321–329
(2000).
18. Z. R. Grabowski, K. Rotkiewicz, W. Rettig, Chem. Rev. 103,
3899–4032 (2003).
19. A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson,
Chem. Rev. 110, 6595–6663 (2010).
20. D. Veldman, S. C. J. Meskers, R. A. J. Janssen, Adv. Funct.
Mater. 19, 1939–1948 (2009).
SUPPLEMENTARY MATERIALS
www.sciencemag.org/content/363/6434/1429/suppl/DC1
Materials and Methods
Supplementary Text
Tables S1 to S7
Figs. S1 to S4
Spectral Data
References (51–65)
5 September 2018; resubmitted 20 November 2018
Accepted 20 February 2019
10.1126/science.aav3200
21. B. Siegmund et al., Nat. Commun. 8, 15421 (2017).
Fu et al., Science 363, 1429–1434 (2019)
29 March 2019
6 of 6

Photocatalytic decarboxylative alkylations mediated by triphenylphosphine and sodium
iodide
Ming-Chen Fu, Rui Shang, Bin Zhao, Bing Wang and Yao Fu
Science 363 (6434), 1429-1434.
DOI: 10.1126/science.aav3200
Crowdsourcing a chromophore
Photoredox catalysis is widely used to accelerate chemical reactions by channeling the energy in visible light.
However, most implementations rely on expensive chromophores to absorb light. Fu et al. now show that a pair of cheap
components acting in concert can induce these reactions, despite not being strong visible absorbers individually. The
combination of sodium iodide and triphenylphosphine allowed photoinduced electron transfer to catalyze a variety of
alkylations.
Science, this issue p. 1429
ARTICLE TOOLS
http://science.sciencemag.org/content/363/6434/1429
SUPPLEMENTARY
MATERIALS
http://science.sciencemag.org/content/suppl/2019/03/27/363.6434.1429.DC1
REFERENCES
This article cites 63 articles, 8 of which you can access for free
http://science.sciencemag.org/content/363/6434/1429#BIBL
PERMISSIONS
http://www.sciencemag.org/help/reprints-and-permissions
Use of this article is subject to the Terms of Service
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. 2017 © The Authors, some rights reserved; exclusive
licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. The title
Science is a registered trademark of AAAS.