MATERIALS CHEMISTRY

Composition-matched molecular“solders” for semiconductorsDmitriy S. Dolzhnikov,1,2 Hao Zhang,1,2 Jaeyoung Jang,1,2 Jae Sung Son,1,2

Matthew G. Panthani,1,2 Tomohiro Shibata,3,4

Soma Chattopadhyay,3,4 Dmitri V. Talapin1,2,5*

We propose a general strategy to synthesize largely unexplored solublechalcogenidometallates of cadmium, lead, and bismuth. These compounds can be usedas “solders” for semiconductors widely used in photovoltaics and thermoelectrics.The addition of solder helped to bond crystal surfaces and link nano- or mesoscaleparticles together. For example, CdSe nanocrystals with Na2Cd2Se3 solder was usedas a soluble precursor for CdSe films with electron mobilities exceeding 300 squarecentimeters per volt-second. CdTe, PbTe, and Bi2Te3 powders were molded into variousshapes in the presence of a small additive of composition-matched chalcogenidometallateor chalcogel, thus opening new design spaces for semiconductor technologies.

Two pieces of metal, such as copper wires,canbemechanically andelectrically connectedby using soldering, a process in which me-tallic items are joined together by introduc-ing a filler metal (solder) with lower melting

point (1). Soldering and related processes (suchas brazing) are very useful in many areas, frommicroelectronics to plumbing. Unlike the casefor metals, there are no established methods forjoining semiconductor pieces under mild condi-tionswithout disrupting semiconducting proper-ties at the joint. A “semiconductor solder” couldaffect numerous fields, including printable elec-tronics, photovoltaics, and thermoelectrics, be-cause it would add new techniques for deviceintegration.The electronic properties of semiconductor

interfaces are much more sensitive to impuritiesand structural defects than are metals. Metal-metal contacts always show ohmic behavior,whereas semiconductor-semiconductor inter-faces are more complex. Misalignment of theFermi energy levels or trapping charge carriersat the interface creates a Schottky barrier (2).An ideal “semiconductor solder” should be con-stituted of a precursor that, upon mild heattreatment, forms a semiconducting materialstructurally and compositionally matched tothe bonded parts. In this study, we designedchalcogenidometallate “solders” for technologi-cally important II-VI, IV-VI, and V-VI semi-conductors, including CdSe, CdTe, HgxCd1−xTe,PbTe, (BixSb1−x)2Te3, and Bi2Te3.The use of chalcogenidometallates as soluble

precursors for inorganic semiconductors was first

introduced by Mitzi et al., who used the abilityof N2H4 to dissolve metal chalcogenides in thepresence of elemental chalcogens (3), formingsoluble hydrazinium chalcogenidometallates.These species cleanly decompose back into semi-conducting metal chalcogenides upon heating.This simple approach works very well for thechalcogenides of Cu, Ga, Ge, In, Sb, Sn, andmixedphases such asCu(In1−xGax)Se2 andCu2ZnSn(S,Se)4(4–6) but showed no success for Cd-, Pb-, and Bi-chalcogenides (7, 8), which are among the mostwidely used binary and ternary semiconductors.Here, we propose the origin of this limitation anda general solution, which substantially expandsthe list of solution-processed semiconductors.In the above process, the formation of a sol-

uble chalcogenidometallate begins with the reduc-tion of an elemental chalcogen (Ch = S, Se, or Te)by N2H4 following Eq. 1 (3, 4)

2nCh + 5N2H4→ 4N2H5+ + 2Chn

2− + N2 (1)

The Chn2− reacts with electron-deficient metal

centers at the surface of a solid metal chalcogen-ide (MxChy), generating soluble MxCh(y+m)

2m−

chalcogenidometallate ions:

2nMxChy + 2mChn2− + 5(n − 1)mN2H4 →

2nMxCh(y+m)2m−+ (n− 1)mN2+4(n− 1)mN2H5

+ (2)

The progression of reaction in Eq. 2 is deter-mined by the balance between the lattice energyof the metal chalcogenide and the free energy offormation and solvation of the chalcogenidome-tallate complex. Reactivity of Chn

2− is highest atn = 1 and decreases as n increases. The reducingpotential of N2H4 in the reaction in Eq. 1 is notsufficient to bring Se and Te into their mostreactive Ch2− state. X-ray absorption near-edgespectroscopy (XANES) measurements show thatEq. 1 generates soluble Se or Te species with anoxidation state near zero, which is equivalent tolarge n (fig. S1) (9). To increase the driving forcefor Eq. 2, we used A2Se or A2Te (A = Na, K, andCs) instead of elemental chalcogens and found

that for n = 1, Eq. 2 proceeds smoothly for a num-ber of metal chalcogenides previously consideredunreactive, including CdSe, CdTe, PbS, PbSe, PbTe,Bi2S3, Bi2Se3, Bi2Te3, (BixSb1−x)2Te3, and some others,which are summarized in (9). These reactions couldalso be conveniently carried out in one pot—for ex-ample, by adding alkali metal hydride to the stoi-chiometricmixture of CdTe andTe inN2H4: CdTe+Te + 2NaH→ Na2CdTe2 + gaseous products (9).The structure of K2CdTe2·2N2H4 crystallized

from the reaction mixture is shown in Fig. 1A. Itcontains molecular chains built of edge-sharing,slightly distorted [CdTe4] tetrahedrons withTe-Cd-Te angles of 99.4° and 119.1° (9). The bondsbetween Cd and Te are 2.81 and 2.83 Å in length,which are similar to that of zinc blende CdTe(2.81Å). This structuralmotifwith one-dimensional(1D) “molecular wires” has apparently not beenreported for chalcogenidocadmates (10, 11). Inthe crystal lattice, [CdTe2

2−]∞ chains are sepa-rated with N2H4 molecules, which leads to thefacile solubility of the compound in various sol-vents, up to 600 mg/mL in N2H4 (Fig. 1B). Thesolubility of ditellurometallates could be furthertailored by means of cation exchange (9). For ex-ample,Na+orK+wasexchanged foralkylammoniumcations such as didodecyldimethyl ammonium(DDA+) or tetraethyl ammonium (NEt4

+), provid-ing good solubility of DDA2CdTe2 in toluene andof (NEt4)2CdTe2 in acetonitrile (CH3CN) or N,N-dimethylfomamide (DMF), respectively (Fig. 1B).Exchange of K+with hydrazinium cations (9) couldbe used to facilitate thermal decomposition ofthe resulting hydrazinium salts.X-ray diffraction and extended x-ray absorp-

tion fine structure (EXAFS) studies (figs S2 andS3 and tables S1 and S2) (9) of ditellurocadmatessuggest that, in solution, [CdTe2

2−]∞ chains canexist in the equilibrium forms outlined in Fig. 1C.In a strongly coordinating solvent such as N2H4,the equilibrium shifted toward the [CdTe(m-Te)]n

2n−

structure, where each Cd atom has 3 Te neigh-bors and Te has on average 1.5 Cd atoms in thefirst coordination shell. On the other hand,weak-ly coordinating solvents such as CH3CN shiftedthe equilibrium toward the structure with twobridging Te atoms per [CdTe2]

2− unit. These var-iations in coordination environmentwere reflectedby fully reversible shifts of the absorption bandsin mixtures of N2H4 and CH3CN with varioussolvent ratios (fig. S4) (9).ReactingNa2CdTe2with a stoichiometric amount

of CdCl2 in N2H4 formed a white amorphous gel,further referred to as “CdTe-gel” (Fig. 1B), whichis an important addition to the chalcogel family(12). The gel was easily separated from its NaClby-product by means of centrifugation and thenredispersed in N-methylformamide (NMF) orN2H4, forming a stable gel of [CdTe2]

2− polyionscross-linkedwith Cd2+. Upon heating above 250°C,the CdTe-gel transformed into crystalline CdTe,as shown in fig. S5 (9).[CdTe2

2−]∞ ions, similar to other chalcogeni-dometallates (13), were used as capping ligandsfor colloidal nanocrystals (NCs). [CdTe2]

2−-cappedCdTe quantum dots showed bright band-edgephotoluminescence (Fig. 1B and fig. S6) (9). The

SCIENCE sciencemag.org 23 JANUARY 2015 • VOL 347 ISSUE 6220 425

1Department of Chemistry, University of Chicago, Chicago, IL60637, USA. 2James Franck Institute, University of Chicago,Chicago, IL 60637, USA. 3Materials Research CollaborativeAccess Team (MRCAT) Sector 10 Advanced Photon Source,Argonne National Laboratory, Argonne, IL 60439, USA.4Department of Physics, Advanced Materials Group, IllinoisInstitute of Technology, Chicago, IL 60616, USA. 5Center forNanoscale Materials, Argonne National Laboratory, Argonne,IL 60439, USA.*Corresponding author. E-mail: dvtalapin@uchicago.edu

RESEARCH | REPORTS

presence of strong emission after the ligand ex-change suggests that [CdTe2]

2− ions did not intro-duce fast recombination centers at the CdTe surface.Following similar approaches, we synthesized

and characterized a series of soluble Cd-, Pb-, andBi chalcogenidometallates and chalcogels (9).These chalcogenidometallates could be dispersed

in polar solvents, includingN2H4, ethylenediamine,NMF, dimethyl sulfoxide (DMSO), formamide (FA),DMF, andN2H4/watermixtures (9). Such chemicalversatility allowed for the design of composi-tionally matched “solders” for practical semi-conductors. These solders were used in variouscapacities—for example, to consolidate nano- or

mesoscopic grains or to bond single-crystal wafers.In one example, we applied Na2Cd2Se3 solder toCdSe NCs: Colloidal 4.7-nm CdSe NCs cappedwith n-octadecylphosphonic acid (ODPA) wereligand-exchanged with [Cd2Se3]

2− ions (9). Theinfrared (IR) absorption spectra confirmed a com-plete replacement of original ODPA ligands

426 23 JANUARY 2015 • VOL 347 ISSUE 6220 sciencemag.org SCIENCE

Fig. 1. Structure and properties of CdTe22– compounds. (A) Crystal structure of K2CdTe2·2N2H4 determined by using single crystal x-ray diffraction.

Hydrogen atoms are not shown for clarity. (B) Solubility of Na2CdTe2 in different solvents can be tuned by cation exchange from sodium (Na+) totetratethylammonium (NEt4

+) or didodecyldimethylammonium (DDA+). Polymeric [CdTe22−]∞ ions can be cross-linked with Cd2+ ions, forming stable “CdTe-gel.”

[CdTe22−]∞ ions can serve as ligands for luminescent CdTe quantumdots (QDs). (C) Depending on solvent polarity and coordinating ability, [CdTe2

2−]∞ can exist intwo equilibrium forms.

Fig. 2. Transport properties ofCdSe NCs after ligandexchange. (A) Absorptionspectra of 4.7-nm CdSe NCscappedwithODPAand [Cd2Se3]

2−

ligands. (Inset) Photograph ofthe solution of [Cd2Se3]

2−-capped CdSe NCs. (B)Scheme of a FETdevice with10-nm ZrO2 gate dielectric andAl source and drain electrodes.(C) Transfer characteristics and(D) output characteristics of anFETwith a channel made ofspin-coated 4.7 nm [Cd2Se3]

2−-capped CdSe NCs annealed at250°C for 30 min. FETchannelwidth and length are 1500 and30 mm, respectively.The inset in(C) shows the linear and satura-tion regime electron mobility extracted from FET transfer characteristics (9).

RESEARCH | REPORTS

with [Cd2Se3]2− ions (fig. S7) (9). A stable col-

loidal solution of [Cd2Se3]2−-capped CdSe NCs

was obtained, which preserved the original quan-tum dot excitonic features, with a small red shiftof the first absorption peak (Fig. 2A). This shiftwas likely caused by the expansion of wave func-tions into the compositionally matched ligandshell. The solution of [Cd2Se3]

2−-capped CdSeNCs was spin-coated on heavily doped Si sub-strates with a ~10-nm-thick ZrOx gate dielectricand annealed under nitrogen at various temper-atures for 30 min. Mobility measurements werecarried out in a standard field-effect transistor(FET) geometry (Fig. 2B) (14, 15). Even annealingat only 250°C was sufficient to achieve excellentoutput and transfer characteristics (Fig. 2, C andD, and fig. S8) (9) with an electron mobility of210 cm2/V s, which exceeds the best reportedvalues for any solution-processed inorganic semi-conductor (table S3) (9). Annealing NC films at300°C results in devices with FET mobility above300 cm2/V s (fig. S9) (9). In control experiments,we found that FETs with channels made of pureCdSe NCs, Na2Cd2Se3, or (N2H5)2Cd2Se3 showedmobilities of <4 cm2/V s (figs. S10 to S12) (9).In recent years, various organic and inorganic

ligands (such as 1,4-phenylenediamine, In2Se42−,

and SCN−) have been used as “electronic linkers”for CdSe NCs (16–20). The highest reported FETmobilities of annealed CdSe NC samples were onthe order of 35 cm2/V s (table S3) (9), which is ~5%of themobility of CdSe single crystals at 300K (21).In our case, [Cd2Se3]

2− ions worked as a compo-sitionally matched solder that improved carriertransport up to almost half of the single-crystal

value, which is surprising given that the Scherrersize of CdSe grains was about 15 and 22 nm forsamples annealed at 250°C and 300°C, respec-tively (fig. S13) (9). This grain size was alsoconfirmed with transmission electron micros-copy (TEM) measurements of annealed films(fig. S14) (9). Various chemical treatments havebeen successfully used to passivate interfacesand control doping in polycrystalline semicon-ductors and NC solids (22, 23). At the same time,it has proven difficult to achieve high mobilitythrough sintering (24) or oriented attachment ofNCs (25). Recent studies show that simply in-creasing CdSe grain size does not improve FETmobility (26). We therefore attribute this highmobility to exceptional electronic transparency ofthe grain boundaries, which did not introduceappreciable transport bottlenecks.The molecular solders functioned not only for

nanoscale but also for mesoscale grains. For ex-ample, PbTe, Bi2Te3, orCdTe filmswere obtainedbysoldering microparticles, obtained by ball-millingof ingots, by using Na2PbTe2, (N2H5)4Bi2Te5, orCdTe-gel as solders, respectively (Fig. 3, B to D,and fig. S15) (9). In the first case, we appliedthe sodium salt because Na+ is a commonly usedp-type dopant for PbTe. A suspension of PbTemicroparticles with a small amount [2 to 5weightpercent (wt %)] of Na2PbTe2 solder was drop-caston a hydrophilized glass substrate and annealedat 500°C for 30 min, resulting in a continuousthin film with a uniform thickness of ~7 mm (9)with well-connected dense grains (Fig. 3B). Thefilm exhibited a Hall mobility of 58 cm2/V s,which is comparable with that (~100 cm2/ V s) of

optimized Na-doped PbTe thermoelectric pelletsmade at a similar temperature bymeans of spark-plasma sintering or hot pressing (27, 28). In con-trol experiments, no continuous filmswereobtainedfrom suspensions of PbTe and other micropar-ticles without addition of solder. After annealingunder identical conditions, the material remainedpowdery (Fig. 3A). The above examples show thatchalcogenidometallates and chalcogels,mixedwithnano- andmicroparticles, offer a viable path towardsolution processing of thin-film semiconductors.The solder effect was used to consolidate mi-

croparticles (such as PbTe, Bi2Te3, Bi0.5Sb1.5Te3,and CdTe) into 3D blocks cast with a mold (Fig.3E). The addition of a few weight percent of mo-lecular or gel solder helped connect the micro-particles and allowed the casting of solid pelletsin graphite molds with no external pressure,similar to molding metal particles (29). Disk-,triangle-, square-, and bar-shaped blocks wereobtained via annealing various combinations ofnano- andmicroparticles (PbTe,Bi2Te3, Bi0.5Sb1.5Te3,or CdTe) with corresponding compositionallymatched molecular chalcogenidometallates orchalcogels (9). In control experiments withoutsolders, the microparticles molded under thesame conditions remained powdery (fig. S16) (9).Comparison of Fig. 3, A and B, shows that theaddition of molecular solder facilitates sinteringand grain growth compared with pure compo-nents. The microscopic origin of this effect islikely related to improved mass transport alongand across the grain boundaries (30).Chalcogenidometallate solders were also used

for bonding metal chalcogenide crystals undermodest pressures and temperatures. Shown inFig. 3F is an example in which two polished (111)CdTe surfaces (with an area of 0.25 cm2) werebonded at ~0.5 MPa and 500°C by using a mi-nuscule amount (2 mL) of Na2CdTe2 solution inN2H4 (9). No physical bonding between CdTecrystals occurred in control experiments withsimilarly prepared crystal surfaces but no solder.The above examples reveal the potential of

bottom-up assembly of inorganic semiconductorswith compositionally matchedmolecular solders.This approach enables the highest electron mo-bility achieved to date for solution-processedsemiconductors, opening new opportunities forprintable electronics and optoelectronics, andpaves the way to future technological develop-ments such as moldable or 3D-printable thermo-electric materials.

REFERENCES AND NOTES

1. Encyclopædia Britannica Online, s. v. “soldering” (2014);available at www.britannica.com/EBchecked/topic/553118/soldering.

2. S. M. Sze, Semiconductor Devices: Physics and Technology(John Wiley and Sons, New York, 2002).

3. D. B. Mitzi, L. L. Kosbar, C. E. Murray, M. Copel, A. Afzali,Nature 428, 299–303 (2004).

4. D. B. Mitzi, Adv. Mater. 21, 3141–3158 (2009).5. D. B. Mitzi et al., Adv. Mater. 20, 3657–3662 (2008).6. T. K. Todorov, K. B. Reuter, D. B. Mitzi, Adv. Mater. 22,

E156–E159 (2010).7. R. Y. Wang et al., Chem. Mater. 22, 1943–1945 (2010).8. D. H. Webber, R. L. Brutchey, J. Am. Chem. Soc. 135,

15722–15725 (2013).

SCIENCE sciencemag.org 23 JANUARY 2015 • VOL 347 ISSUE 6220 427

Fig. 3. Soldering variousmaterials with novel chalcogenidometallates.Scanning electronmicroscopyimages of (A) ball-milled PbTe microparticles annealed at 500°C for 30 min. (B) Film of similar PbTemicroparticles annealed in the presence of 5 wt % Na2PbTe2. (C) Film made of Bi2Te3 microparticlesannealed in the presence of 5 wt % (N2H5)4Bi2Te5. (D) Filmmade of CdTe NCs annealed in the presence of20wt% “CdTe-gel.”The filmswere annealed at 500°C,400°C, and 450°C for 30min, respectively. (Insets)Film optical micrographs. (E) 3D blocks of PbTe, Bi2Te3, and CdTe cast by using graphite molds of variousshapes frommicroparticles mixed with corresponding “solders”: Na2PbTe2 for PbTe, Na4Bi2Te5 for Bi2Te3,and Na2CdTe2 for CdTe. (F) A photograph showing two CdTe crystals with mechanically polished (111)faces soldered with Na2CdTe2 at 500°C.

RESEARCH | REPORTS

9. Materials and methods are available as supplementarymaterials on Science Online.

10. R. C. Burns, L. A. Devereux, P. Granger, G. J. Schrobilgen,Inorg. Chem. 24, 2615–2624 (1985).

11. C.-W. Park, R. J. Salm, J. A. Ibers, Can. J. Chem. 73, 1148–1156(1995).

12. S. Bag, P. N. Trikalitis, P. J. Chupas, G. S. Armatas,M. G. Kanatzidis, Science 317, 490–493 (2007).

13. M. V. Kovalenko, M. Scheele, D. V. Talapin, Science 324,1417–1420 (2009).

14. D. V. Talapin, C. B. Murray, Science 310, 86–89 (2005).15. D. S. Chung et al., Nano Lett. 12, 1813–1820 (2012).16. Y. Liu et al., Nano Lett. 10, 1960–1969 (2010).17. Y. Liu et al., Nano Lett. 13, 1578–1587 (2013).18. J. H. Choi et al., Nano Lett. 12, 2631–2638 (2012).19. J. S. Lee, M. V. Kovalenko, J. Huang, D. S. Chung, D. V. Talapin,

Nat. Nanotechnol. 6, 348–352 (2011).20. D. Yu, C. Wang, P. Guyot-Sionnest, Science 300, 1277–1280 (2003).21. S. S. Devlin, Physics and Chemistry of II-VI Compounds

(North Holland, Amsterdam, 1967).

22. B. M. Basol, S. S. Ou, O. M. Stafsudd, J. Appl. Phys. 58,3809–3813 (1985).

23. S. J. Oh et al., Nano Lett. 14, 1559–1566 (2014).24. B. A. Ridley, B. Nivi, J. M. Jacobson, Science 286, 746–749

(1999).25. W. J. Baumgardner, K. Whitham, T. Hanrath, Nano Lett. 13,

3225–3231 (2013).26. Z. M. Norman, N. C. Anderson, J. S. Owen, ACS Nano 8,

7513–7521 (2014).27. S. N. Girard et al., J. Am. Chem. Soc. 133, 16588–16597

(2011).28. K. Biswas et al., Nature 489, 414–418 (2012).29. R. Klajn et al., Science 316, 261–264 (2007).30. S.-J. L. Kang, Sintering: Densification, Grain Growth and

Microstructure (Butterworth-Heinemann, Oxford, 2004).

ACKNOWLEDGMENTS

We thank S. Kwon and V. Zyryanov for help with EXAFSmeasurements and N. James for reading the manuscript. Thework was supported by the II-VI Foundation, U.S. Department of

Energy (DOE) SunShot program under award DE-EE0005312 andby NSF under award DMR-1310398. D.V.T. also thanks the KeckFoundation. This work used facilities supported by the NSFMaterials Research Science and Engineering Center Programunder awards DMR 08-20054 and DMR 14-20709. Use of theCenter for Nanoscale Materials and Advanced Photon Source wassupported by the DOE, Office of Science, Office of Basic EnergySciences, under contract DE-AC02-06CH11357. MRCAT operationsare supported by the DOE and MRCAT host institutions.

SUPPORTING ONLINE MATERIAL

www.sciencemag.org/content/347/6220/425/suppl/DC1Materials and MethodsFigs. S1 to S16Tables S1 to S3References (31–48)

27 August 2014; accepted 8 December 201410.1126/science.1260501

DRUG RESISTANCE

K13-propeller mutations conferartemisinin resistance in Plasmodiumfalciparum clinical isolatesJudith Straimer,1 Nina F. Gnädig,1 Benoit Witkowski,2* Chanaki Amaratunga,3*Valentine Duru,2* Arba Pramundita Ramadani,4,5*† Mélanie Dacheux,1

Nimol Khim,2 Lei Zhang,6 Stephen Lam,6 Philip D. Gregory,6 Fyodor D. Urnov,6

Odile Mercereau-Puijalon,7 Françoise Benoit-Vical,4,5‡ Rick M. Fairhurst,3‡Didier Ménard,2‡ David A. Fidock1,8§

The emergence of artemisinin resistance in Southeast Asia imperils efforts to reducethe global malaria burden. We genetically modified the Plasmodium falciparum K13 locususing zinc-finger nucleases and measured ring-stage survival rates after drug exposurein vitro; these rates correlate with parasite clearance half-lives in artemisinin-treatedpatients. With isolates from Cambodia, where resistance first emerged, survival ratesdecreased from 13 to 49% to 0.3 to 2.4% after the removal of K13 mutations. Conversely,survival rates in wild-type parasites increased from ≤0.6% to 2 to 29% after the insertionof K13 mutations. These mutations conferred elevated resistance to recent Cambodianisolates compared with that of reference lines, suggesting a contemporary contributionof additional genetic factors. Our data provide a conclusive rationale for worldwideK13-propeller sequencing to identify and eliminate artemisinin-resistant parasites.

The worldwide use of artemisinin (ART)–based combination therapies (ACTs) forthe treatment of Plasmodium falciparummalaria is the foundation of renewed effortsto eradicate this leading cause of childhood

mortality (1, 2). The pharmacodynamic proper-ties of clinically usedART derivatives [artesunate,artemether, and dihydroartemisinin (DHA)] canreduce the biomass of drug-sensitive parasitesby four orders of magnitude every 48 hours (3),corresponding to a single cycle of asexual blood-stage P. falciparum development. The shorthalf-life (typically <1 hour) of ART derivatives inplasma necessitates the use of longer-lastingpartner drugs that can eliminate residual par-asites once the ART component has dropped tosubtherapeutic concentrations (4). The use ofACTs in expanded malaria control and elimina-tion programs has yielded notable successes inrecent years, contributing to an estimated 30%

reduction in global mortality rates in the pastdecade (5).These impressive gains, however, are now

threatened by the emergence of ART resistance,first detected in western Cambodia and now ob-served in Thailand, Vietnam, andMyanmar (6, 7).The severity of this situation is underscored bythe fact that resistance to piperaquine, an ACTpartner drug, is emerging in western Cambodia(8, 9). No alternative, fully effective first-line ther-apy is currently available to replace ACTs, shouldART fail globally. Clinically, ART resistance is de-fined as a long parasite clearance half-life (thetime it takes for the peripheral blood parasitedensity to decrease by 50%) after treatment withARTmonotherapy or anACT (6, 10, 11). Thismetriccorrelates with the percentage of early “ring-stage”parasites (0 to 3 hours after invasion of humanerythrocytes) that survive a pharmacologicallyrelevant exposure to DHA (the active metabolite

of all ARTs), asmeasured in the in vitro Ring-stageSurvival Assay (RSA0-3h) (12).Recently, mutations in the propeller domain

of the K13 gene were identified as candidate mo-lecular markers of ART resistance (13). This generesides on chromosome 13 of the P. falciparumgenome, near regions earlier associatedwith slowparasite clearance rates (14–16). K13 belongs tothe kelch superfamily of proteins, whose propellerdomain harbors multiple protein-protein interac-tion sites andmediates diverse cellular functions,including ubiquitin-regulated protein degradationand oxidative stress responses (17). The K13M476Imutation was first observed in Tanzanian F32parasites that were exposed in vitro to escalatingconcentrations of ART over 5 years, yieldingthe F32-ART line (13, 18). [Single-letter abbrevia-tions for the amino acid residues are as follows:A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly;H,His;I, Ile; K, Lys; L, Leu; M,Met; N, Asn; P, Pro; Q, Gln;R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.In themutants, other amino acidswere substitutedat certain locations; for example,M476I indicatesthat methionine at position 476 was replacedby isoleucine.] Subsequent genomic analysis ofCambodian isolates identified four prevalentK13-propeller mutations (Y493H, R539T, I543T,and C580Y) that were associated with elevatedRSA0-3h survival rates in vitro and long parasiteclearance half-lives (>5 hours) in patients (13, 19).

428 23 JANUARY 2015 • VOL 347 ISSUE 6220 sciencemag.org SCIENCE

1Department of Microbiology and Immunology, ColumbiaUniversity College of Physicians and Surgeons, New York,NY, USA. 2Malaria Molecular Epidemiology Unit, InstitutPasteur du Cambodge, Phnom Penh, Cambodia. 3Laboratoryof Malaria and Vector Research, National Institute of Allergyand Infectious Diseases, National Institutes of Health,Bethesda, MD, USA. 4Centre National de la RechercheScientifique (CNRS), Laboratoire de Chimie de CoordinationUPR8241, Toulouse, France. 5Université de Toulouse, UPS,Institut National Polytechnique de Toulouse, Toulouse,France. 6Sangamo BioSciences, Richmond, CA, USA. 7InstitutPasteur, Parasite Molecular Immunology Unit, Paris, France.8Division of Infectious Diseases, Department of Medicine,Columbia University College of Physicians and Surgeons,New York, NY, USA.*These authors contributed equally to this work. †Present address:Department of Pharmacology and Therapy, Faculty of Medicine,Gadjah Mada University, Yogyakarta, Indonesia. ‡These authorscontributed equally to this work. §To whom correspondenceshould be addressed. E-mail: df2260@columbia.edu

RESEARCH | REPORTS