REFERENCES AND NOTES

1. O. M. Yaghi et al., Nature 423, 705–714 (2003).2. L. Carlucci, G. Ciani, D. M. Proserpio, Coord. Chem. Rev. 246,

247–289 (2003).3. B. Chen, M. Eddaoudi, S. T. Hyde, M. O’Keeffe, O. M. Yaghi,

Science 291, 1021–1023 (2001).4. V. A. Blatov, L. Carlucci, G. Ciani, D. M. Proserpio, Cryst. Eng.

Comm. 6, 378–395 (2004).5. S. R. Batten, R. Robson, Angew. Chem. Int. Ed. 37, 1460–1494

(1998).6. T. K. Maji, R. Matsuda, S. Kitagawa, Nat. Mater. 6, 142–148

(2007).7. C. O. Dietrich-Buchecker, J.-P. Sauvage, Angew. Chem. Int. Ed. Engl.

28, 189–192 (1989).8. C. O. Dietrich-Buchecker, J.-P. Sauvage, J. M. Kern, J. Am.

Chem. Soc. 106, 3043–3045 (1984).9. M.-C. Jiménez, C. Dietrich-Buchecker, J.-P. Sauvage,

Angew. Chem. Int. Ed. 39, 3284–3287 (2000).10. S. V. Pakhomova, M. A. Proskurnin, V. V. Chernysh,

M. Y. Kononets, E. K. Ivanova, J. Anal. Chem. 56, 910–917(2001).

11. M. Linke et al., J. Am. Chem. Soc. 122, 11834–11844(2000).

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

13. M. T. Miller, P. K. Gantzel, T. B. Karpishin, Inorg. Chem. 37,2285–2290 (1998).

14. Y.-B. Zhang et al., J. Am. Chem. Soc. 135, 16336–16339(2013).

15. F. J. Uribe-Romo et al., J. Am. Chem. Soc. 131, 4570–4571(2009).

16. X. Wu, K. W. Zilm, J. Magn. Reson. 102, 205–213(1993).

17. M. Gemmi, P. Oleynikov, Z. Kristallogr. 228, 51–58(2013).

18. E. Mugnaioli et al., Angew. Chem. Int. Ed. 51, 7041–7045(2012).

19. Q. Sun et al., J. Mater. Chem. A 2, 17828–17839(2014).

20. E. V. Alexandrov, V. A. Blatov, D. M. Proserpio, Acta Crystallogr.A 68, 484–493 (2012).

21. D. V. Scaltrito, D. W. Thompson, J. A. O’Callaghan, G. J. Meyer,Coord. Chem. Rev. 208, 243–266 (2000).

22. W. C. Oliver, G. M. Pharr, J. Mater. Res. 19, 3–20(2004).

23. J. C. Tan, A. K. Cheetham, Chem. Soc. Rev. 40, 1059–1080(2011).

ACKNOWLEDGMENTS

The structures of COF-505 and Cu(PDB)2(BF4) are availablefree of charge from the Cambridge Crystallographic DataCentre under the reference numbers CCDC-1434851 andCCDC-1434852, respectively. This research was supported byBASF SE (Ludwigshafen, Germany) for synthesis and basiccharacterization, and the U.S. Department of Defense, DefenseThreat Reduction Agency (HDTRA 1-12-1-0053) for mechanicalproperties. We thank C. Canlas for his assistance with solid-stateNMR and A. Schöedel (Yaghi group), B. Zhang, and Y. Liu(Molecular Foundry, Lawrence Berkeley National Laboratory)for helpful discussions. This work was also supported by theSpanish Ministry of Economy and Competitiveness through theJuan de la Cierva program (F.G.); a Grant-in-Aid for ScientificResearch (C) (25390023) and JST Research Acceleration Program(Z.L. and K.S.); grants from Vetenskapsrådet (Y.M. and P.O.)and JEOL Ltd (P.O.), Japan; EXSELENT and 3DEM-Natur, Sweden(O.T.); and BK21Plus, Korea (O.T.). Beamline 7.3.3 of theAdvanced Light Source is supported by the Director of theOffice of Science, Office of Basic Energy Sciences, of the U.S.Department of Energy under contract DE-AC02-05CH11231. TheAFM study was supported by the National Science Foundation(NSF) (grant DMR‐1344290). The data reported in the paper arepresented in the supplementary materials.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/351/6271/365/suppl/DC1Materials and MethodsFigs. S1 to S21Tables S1 to S4Reference (24)

8 September 2015; accepted 25 November 201510.1126/science.aad4011

PHOTOPHYSICS

Direct observation of tripletenergy transfer fromsemiconductor nanocrystalsCédric Mongin,1 Sofia Garakyaraghi,1 Natalia Razgoniaeva,2

Mikhail Zamkov,2 Felix N. Castellano1*

Triplet excitons are pervasive in both organic and inorganic semiconductors but generallyremain confined to the material in which they originate. We demonstrated by transientabsorption spectroscopy that cadmium selenide semiconductor nanoparticles, selectivelyexcited by green light, engage in interfacial Dexter-like triplet-triplet energy transfer withsurface-anchored polyaromatic carboxylic acid acceptors, extending the excited-statelifetime by six orders of magnitude. Net triplet energy transfer also occurs from surfaceacceptors to freely diffusing molecular solutes, further extending the lifetime whilesensitizing singlet oxygen in an aerated solution. The successful translation of tripletexcitons from semiconductor nanoparticles to the bulk solution implies that suchmaterials are generally effective surrogates for molecular triplets. The nanoparticles couldthereby potentially sensitize a range of chemical transformations that are relevant forfields as diverse as optoelectronics, solar energy conversion, and photobiology.

Semiconductor nanocrystals represent an im-portant class of stable light-emitting mate-rials that can be systematically tuned as aresult of size-dependent quantum confine-ment, producing intense absorptions and

photoluminescence ranging from the ultraviolet(UV) to the near-infrared (near-IR) (1, 2). Theirprominence continues to expand, owing to ex-tensive optoelectronic, photochemical, and bio-medical applications (3–9). Substantial researcheffort has been expended on funneling energyinto these nanomaterials to produce enhancedphotoluminescence via Förster transfer and onexploiting the energized semiconductor nano-crystals to deliver or accept electrons from sub-strates (10–14), sometimes en route to solar fuelsphotosynthesis (15–18). Tabachnyk et al. andThompson et al. independently demonstratedthe reverse triplet energy transfer process to thatdescribed here, wherein molecular organic semi-conductors transfer their triplet energy to PbSeand PbS nanocrystals in thin films that interfaceboth materials (19, 20). However, the extractionof triplet excitons from semiconductor quan-tum dots and related inorganic nanomaterialsremains largely unexplored. Semiconductor nano-crystals potentially offer considerable advantagesover molecular photosensitizers in terms of facilepreparative synthesis, photostability, size-tunableelectronic and photophysical properties, highmolar extinction coefficients, and trivial postsyn-thesis functionalization. Moreover, the inherentlylarge (and energy-consuming) singlet-triplet en-ergy gaps characteristic of molecular sensitizers

can be circumvented by using nanomaterials withill-defined spin quantum numbers and closelyspaced (1 to 15 meV) excited-state energy levels(21–24). The broadband light-absorption prop-erties of inorganic semiconductors are extend-able into the near-IR region and can potentiallybe exploited for numerous triplet excited-statereactions, thus enabling stereoselective photo-chemical synthesis, photoredox catalysis, singletoxygen generation, photochemical upconversion,and excited-state electron transfer. Here we pro-vide definitive experimental evidence that tripletenergy transfer proceeds rapidly and efficientlyfrom energized semiconductor nanocrystals tosurface-anchored molecular acceptors. Specifical-ly, CdSe nanocrystals are shown to serve as ef-fective surrogates for molecular triplet sensitizersand can readily transfer their triplet excitons toorganic acceptors at the interface with near-quantitative efficiency.The nanoparticle-to-solution triplet exciton trans-

fer strategy that we implemented is shown sche-matically in Fig. 1; this diagram depicts all of therelevant photophysical processes and the associ-ated energy levels promotingmaterial-to-moleculetriplet excitonmigration.We employed oleic acid(OA)–capped CdSe nanocrystals (CdSe-OA) as thelight-absorbing triplet sensitizer in conjunctionwith 9-anthracenecarboxylic acid (ACA) and 1-pyrenecarboxylic acid (PCA) as triplet acceptorsin toluene. The carboxylic acid functionality en-ables adsorption of these chromophores on theCdSe surface through displacement of the OAcapping ligands; subsequent washing steps iso-late the desired CdSe/ACA or CdSe/PCA donor/acceptor systems. Selective green light excita-tion of CdSe/ACA or CdSe/PCA sensitizes trip-let exciton migration from the semiconductorto the surface-bound molecular acceptor. Wedirectly visualized this interfacial Dexter-like

SCIENCE sciencemag.org 22 JANUARY 2016 • VOL 351 ISSUE 6271 369

1Department of Chemistry, North Carolina State University,Raleigh, NC 27695-8204, USA. 2Department of Physics andCenter for Photochemical Sciences, Bowling Green StateUniversity, Bowling Green, OH 43403, USA.*Corresponding author. E-mail: fncastel@ncsu.edu

RESEARCH | REPORTSon June 3, 2020

http://science.sciencemag.org/

Dow

nloaded from

triplet-triplet energy transfer (TTET) by moni-toring the kinetic growth of the characteristictriplet-to-triplet (T1 → Tn) absorptions in ACAand PCA (wavelength of maximum absorbance =430 nm) using ultrafast transient absorption(TA) spectroscopy (25, 26). The long-lived ACAand PCA localized triplets furthermore enabledexothermic triplet energy transfer to freely diffus-ing 2-chlorobisphenylethynylanthracene (CBPEA)and dioxygen.CdSe-OA suspended in toluene was prepared

as described in the supplementarymaterials. Thefirst exciton band in these samples was locatedat 505 nm (2.46 eV; molar extinction coefficientat 505 nm, 59,200 M−1 cm−1); using an estab-lished empirical equation (27), the average di-ameter of these nanoparticles was estimated tobe 2.4 nm, in good agreement with transmissionelectron microscopy results (fig. S1). The CdSe-OA photoluminescence features, including thespectrally narrow band-gap “bright state” and thelower-energy “trap state” emissions, are shown inFig. 2A (28). Triplet excitons derived from theseexcited states (2.40 eV and 2.30 to 1.40 eV, re-spectively) are suitable for exothermic TTET toACA (lowest triplet state energy, ET = 1.83 eV)and PCA (ET = 2.00 eV) (25, 29). We observedno evidence of triplet energy transfer from theCdSe-OA nanoparticles to anthracene (ET =1.85 eV), 9,10-diphenylanthracene (ET = 1.77 eV),or pyrene (ET = 2.10 eV) acceptors (29), despitethermodynamic expectations that such trans-fers should be exothermic. Unlike ACA and PCA,these acceptors lack the carboxylate functionalityto coordinate directly to the nanoparticle surface.In these cases, the putative bimolecular transferrate through the intervening OA layer appears tobe slower than the excited-state lifetime of thenanoparticle.Photoluminescence from CdSe-OA was quan-

titatively quenched by ACA and PCA (fig. S5). Thenanocrystals bearing surface-anchored molecu-lar acceptors were purified by successive precip-itation and centrifugation cycles, after which thefinal ratio of acceptor to CdSe was determined tobe ~12:1 (fig. S6). Ultrafast TA experiments wereperformed on the quantitatively quenched CdSe-OA/ACA materials to establish the mechanismand time scale for the semiconductor-to-moleculetriplet exciton transfer process. Control experi-ments with CdSe-OA in toluene were also per-formed in the absence of ACA (Fig. 2B). In allcases, symmetric decay of the transient signalwas observed over the first few picoseconds,which is consistent with multi-exciton annihi-lation within the nanocrystals (30), and was con-firmed by laser power dependence experiments(fig. S7). With surface-anchored ACA (Fig. 2C),decay of the CdSe excited state was observed with-in 2 ns, coinciding with the growth of an absorp-tion band centered at 433 nm,whichwas assignedto the T1 ⟶ Tn transitions of ACA (25). Theseresults confirm direct TTET (no intermediates)from the selectively excited CdSe nanocrystalsto the surface-anchored ACA chromophores.The possibility of an electron transfer mecha-nism was eliminated, based on the absence of

370 22 JANUARY 2016 • VOL 351 ISSUE 6271 sciencemag.org SCIENCE

Fig. 1. Illustration of nanocrystal-to-solution triplet energy transfer, the associated energy levels,and the various TTETand decay pathways investigated in this study. PDT, photodynamic therapy.

Fig. 2. Ultrafast spectroscopic evidence for triplet energy transfer from optically excited CdSe nano-crystals to surface-bound ACA. (A) Normalized electronic absorption (solid lines) and emission spectra(dashed lines) of ACA (blue) and CdSe nanocrystals (green) in toluene (OD, optical density). (B andC) Ultrafast TA difference spectra of CdSe-OA nanocrystals in toluene solution upon selective excitation ofCdSe, using 500-nmpulsed laser excitation [0.05 mJ per pulse, 100 fs full width at halfmaximum(FWHM)],in (B) the absence and (C) the presence of surface-anchored ACA in toluene (DA, change in absorbance;hki, average rate constant).The inset in (C) shows TA kineticsmonitored for the growth of 3ACA at 441 nm.(D) Ground-state recovery of CdSemonitored by kinetics at 490 nm, illustrating quantitative quenching ofCdSe in the presence of surface-anchored ACA. Complementary data for PCA are shown in figs. S2 to S4.

RESEARCH | REPORTSon June 3, 2020

http://science.sciencemag.org/

Dow

nloaded from

the ACA radical cation band that would be ex-pected near 750 nm (figs. S8 to 11A) (31). Singlet-singlet energy transfer is not thermodynamicallyfavorable, and no transient signals correspond-ing to singlet ACA (1ACA*) were observed. Tran-sient kinetics were monitored at 441 nm, wherean isosbestic point was present in the CdSe-OAcontrol difference spectra (Fig. 2B). In the pres-ence of ACA (Fig. 2C), this isosbestic point shiftsbecause of the overlapping ACA T1 → Tn absorp-tion band (fig. S12); thus, any new absorptionfeature observed at 441 nm (Fig. 2C, inset) can beattributed to triplet ACA (3ACA*). The quantumefficiency for TTET from CdSe-OA to surface-anchored ACA was determined to be 0.92 fromTA experiments (fig. S12B). Therefore, most ofthe excited states produced within CdSe can beextracted as long-livedmolecular triplets that aresuitable for subsequent chemical reactions. Var-iable rates of energy transfer were anticipated,depending on the initially populated CdSe ex-cited states. Accordingly, following a publishedprecedent (11, 32), a stretched exponential func-tionwas determined to bestmodel the interfacialTTET dynamics (eqs. S1 to S3). Kinetic analysis at441 nm revealed a rise time associated with the3ACA* absorption with an average rate constantof 2.2 × 109 s−1. Additionally, the rise of 3ACA*wascorrelated to the ground-state recovery of CdSe(TA probe wavelength: 480, 490, and 510 nm)(Fig. 2D and fig. S13). Rate constants between2.0 × 109 and 2.8 × 109 s−1 (table S1) were mea-sured, in good agreement with the formation rateof 3ACA*. TTET also occurred, with similar effi-ciencies but slower kinetics, when PCA was usedas the surface-bound molecular acceptor (figs.S3 to S5). This is consistent with the lower TTETdriving force associated with that process.The T1→ Tn absorption centered near 430 nm

that was observed in the ultrafast TA exper-iments (Fig. 2C), which is characteristic of 3ACA*,appeared as a prompt signal in nanosecond flashphotolysis (Fig. 3A). Complementary data forPCA show similar results and are presented inFig. 3B. The transient signals of the acceptorT1 → Tn decay to the ground state over the nextseveral milliseconds, illustrating that excitonscan indeed be harvested from these semiconduc-tor nanomaterials, resulting in an excited-statelifetime enhancement of six orders ofmagnitude.The CdSe/ACA and CdSe/PCA materials also

sensitized the production of 1O2* when the solu-tions were aerated, evidenced by its character-istic photoluminescence centered at 1277 nm inthe near-IR region (fig. S14). This is a uniqueexample of 1O2* sensitization through metal chal-cogenide nanocrystals enabled by a mechanismdistinct from that of Förster transfer (33–35). Tofurther illustrate that excitons could be trans-ferred away from the CdSe interface into thebulk solution, a secondmolecular triplet accep-tor (CBPEA) was added to the CdSe/acceptorsolution. CBPEA possesses a low-lying tripletstate (ET < 1.61 eV), facilitating exothermic TTETfrom energized acceptor chromophores, and isreadily discernible by its distinct T1 → Tn excited-state absorption, centeredat 490nm(36). Figure 3C

presents the TA difference spectra measured at3 ms and 200 ms after 505-nm pulsed nanosecondlaser excitation of the CdSe/ACA solution in thepresence of 6 mM CBPEA (Fig. 3D shows thecomplementary data for PCA as the acceptor). Atearly delay times, the 3ACA* absorption spec-trum dominates (Fig. 3C, blue line), eventuallygiving way quantitatively (quantum efficiency forTTET ∼ 1.0) to a spectrum characteristic of tripletCBPEA (3CBPEA*) at longer delay times (Fig. 3C,red line). By 3ms after the laser pulse, the samplecompletely relaxed to the ground state (Fig. 3C,green line). These data demonstrate that the for-mation of 3ACA* on the CdSe surface is followedbydiffusion-controlled triplet-triplet energy trans-fer to CBPEA (fig. S15; rate constant for energytransfer, 1.2 × 109 M−1 s−1), effectively transferringthe triplet exciton into the bulk solution. Thedynamics of this energymigrationwere capturedin the single-wavelength absorption transientsmeasured at 430 and 490 nm after the 505-nmexcitation pulse (Fig. 3C, inset). The promptly

formed transient absorption signal at 430 nmfrom 3ACA* (red squares) exhibits an excited-statedecay that is kinetically correlated with the forma-tion of 3CBPEA* (blue circles) at 490 nm. Theseexperiments illustrate that bimolecular excited-state chemistry readily proceeds from the CdSe/acceptor materials, exhibiting behavior character-istic of a donor/acceptor TTET molecular system.The results presented here provide proof-of-

concept that excitons can be extracted from thisparticular semiconductor through direct TTET,but it stands to reason that this general strategy isprobably also applicable to a plethora of associ-ated materials. In this regard, molecular triplet-triplet annihilation processes can be sensitized byenergized semiconductor nanocrystals (23, 37, 38).Although this investigation specifically targetedmechanistic insights into the TTET process alongwith chemically relevant triplet exciton decay path-ways, we expect related photochemistry to pro-mote similar triplet energy transfer phenomenain solid-state optoelectronic devices.

SCIENCE sciencemag.org 22 JANUARY 2016 • VOL 351 ISSUE 6271 371

Fig. 3. Kinetic profiles and quenching studies of ACA and PCA triplet states populated fromexcited CdSe nanocrystals. TA difference spectra of a toluene solution of (A) CdSe/ACA (8 mM)measured from 2 ms to 5 ms and (B) CdSe/PCA (8 mM) measured from 2 ms to 10 ms after a 505-nmlaser pulse (1mJ, 5 to 7 ns FWHM; hti, average lifetime).The insets showTAdecay kinetics at 430 nm (graysquares) and their respective fits to eq. S1, illustrating the triplet decay. (C and D) TA difference spectra(excitation wavelength, 505 nm; 5 to 7 ns FWHM; 1 mJ) measured at selected delay times after the laserpulse in (C) CdSe/ACA (5 mM) and CBPEA (6 mM) and (D) CdSe/PCA (5 mM) and CBPEA (6 mM) indeaearated toluene at room temperature.The insets show TA decay kinetics at 430 nm (red circles) andthe rise and decay at 490 nm (blue squares), with their respective biexponential fit lines (solid anddashed), illustrating the triplet energy transfer reaction between 3ACA* and CBPEA.

RESEARCH | REPORTSon June 3, 2020

http://science.sciencemag.org/

Dow

nloaded from

REFERENCES AND NOTES

1. C. B. Murray, D. J. Norris, M. G. Bawendi, J. Am. Chem. Soc.115, 8706–8715 (1993).

2. M. L. Landry, T. E. Morrell, T. K. Karagounis, C.-H. Hsia,C.-Y. Wang, J. Chem. Educ. 91, 274–279 (2014).

3. I. L. Medintz, H. T. Uyeda, E. R. Goldman, H. Mattoussi, Nat.Mater. 4, 435–446 (2005).

4. M. Bruchez Jr., M. Moronne, P. Gin, S. Weiss, A. P. Alivisatos,Science 281, 2013–2016 (1998).

5. I. Gur, N. A. Fromer, M. L. Geier, A. P. Alivisatos, Science 310,462–465 (2005).

6. W. C. W. Chan, S. Nie, Science 281, 2016–2018 (1998).7. S. Coe, W.-K. Woo, M. Bawendi, V. Bulović, Nature 420,

800–803 (2002).8. V. L. Colvin, M. C. Schlamp, A. P. Alivisatos, Nature 370,

354–357 (1994).9. S. A. McDonald et al., Nat. Mater. 4, 138–142 (2005).10. C. Burda, T. C. Green, S. Link, M. A. El-Sayed, J. Phys. Chem. B

103, 1783–1788 (1999).11. I. Robel, M. Kuno, P. V. Kamat, J. Am. Chem. Soc. 129,

4136–4137 (2007).12. K. E. Knowles, M. Malicki, E. A. Weiss, J. Am. Chem. Soc. 134,

12470–12473 (2012).13. R. Reisfeld, M. Gaft, T. Saridarov, G. Panczer, M. Zelner, Mater.

Lett. 45, 154–156 (2000).14. M. Wojdak et al., Phys. Rev. B 69, 233315 (2004).15. Z. Han, F. Qiu, R. Eisenberg, P. L. Holland, T. D. Krauss, Science

338, 1321–1324 (2012).16. K. Wu, Q. Li, Y. Du, Z. Chen, T. Lian, Chem. Sci. 6, 1049–1054

(2015).17. H. B. Yang et al., ACS Nano 8, 10403–10413 (2014).18. E. Khon et al., Nano Lett. 13, 2016–2023 (2013).19. M. Tabachnyk et al., Nat. Mater. 13, 1033–1038 (2014).20. N. J. Thompson et al., Nat. Mater. 13, 1039–1043

(2014).21. G. D. Scholes, Adv. Funct. Mater. 18, 1157–1172 (2008).22. A. L. Efros et al., Phys. Rev. B 54, 4843–4856 (1996).23. M. Wu et al., Nat. Photon. 10, 31–34 (2016).24. S. A. Crooker, T. Barrick, J. A. Hollingsworth, V. I. Klimov, Appl.

Phys. Lett. 82, 2793 (2003).25. S. Hirayama, J. Chem. Soc., Faraday Trans. I 78, 2411

(1982).26. Y. Niko, Y. Hiroshige, S. Kawauchi, G. Konishi, Tetrahedron 68,

6177–6185 (2012).27. W. W. Yu, L. Qu, W. Guo, X. Peng, Chem. Mater. 15, 2854–2860

(2003).28. D. R. Baker, P. V. Kamat, Langmuir 26, 11272–11276

(2010).29. M. Montalti, A. Credi, L. Prodi, T. Gandolfi, Handbook of

Photochemistry (CRC Press, Boca Raton, FL, ed. 3, 2006).30. H. Zhu, Y. Yang, T. Lian, Acc. Chem. Res. 46, 1270–1279

(2013).31. J. M. Masnovi, E. A. Seddon, J. K. Kochi, Can. J. Chem. 62,

2552–2559 (1984).32. V. N. Soloviev, A. Eichhöfer, D. Fenske, U. Banin, J. Am. Chem.

Soc. 123, 2354–2364 (2001).33. R. Bakalova, H. Ohba, Z. Zhelev, M. Ishikawa, Y. Baba, Nat.

Biotechnol. 22, 1360–1361 (2004).34. E. I. Zenkevich et al., J. Phys. Chem. C 115, 21535–21545

(2011).35. A. C. S. Samia, X. Chen, C. Burda, J. Am. Chem. Soc. 125,

15736–15737 (2003).36. T. N. Singh-Rachford, F. N. Castellano, Inorg. Chem. 48,

2541–2548 (2009).37. Z. Huang et al., Nano Lett. 15, 5552–5557 (2015).38. Z. Huang et al., Chem. Mater. 27, 7503–7507 (2015).

ACKNOWLEDGMENTS

This work was supported by the Air Force Office of ScientificResearch (grant FA9550-13-1-0106) and the Ultrafast Initiative ofthe U. S. Department of Energy, Office of Science, Office of BasicEnergy Sciences, through Argonne National Laboratory undercontract no. DE-AC02-06CH11357. M.Z. was supported by NSF(grant CHE-1465052).

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/351/6271/369/suppl/DC1Materials and MethodsFigs. S1 to S15Table S1References (39–41)

14 October 2015; accepted 9 December 201510.1126/science.aad6378

GEOCHEMISTRY

Archean upper crust transitionfrom mafic to felsic marks theonset of plate tectonicsMing Tang,1* Kang Chen,2,1 Roberta L. Rudnick1†

The Archean Eon witnessed the production of early continental crust, the emergence oflife, and fundamental changes to the atmosphere. The nature of the first continental crust,which was the interface between the surface and deep Earth, has been obscured by theweathering, erosion, and tectonism that followed its formation. We used Ni/Co and Cr/Znratios in Archean terrigenous sedimentary rocks and Archean igneous/metaigneous rocksto track the bulk MgO composition of the Archean upper continental crust. This crustevolved from a highly mafic bulk composition before 3.0 billion years ago to a felsic bulkcomposition by 2.5 billion years ago. This compositional change was attended by a fivefoldincrease in the mass of the upper continental crust due to addition of granitic rocks,suggesting the onset of global plate tectonics at ~3.0 billion years ago.

Magnesium content (and its ratio to otherelements) is commonly used as an indexof igneous differentiation and meltingconditions, which are responsible formuch of the compositional variation seen

in silicate rocks. Thus, MgO content serves as afirst-order measure of silicate rock differentiation.Estimating the average MgO content in the uppercontinental crust, and from that the bulk com-position of this crust is, however, challenging.There are two basic approaches to determine thecomposition of the upper continental crust (1–3):(i) weighted averages of surface rocks and (ii)average compositions of terrigneous sedimentssuch as shales (1, 2) and glacial diamictites (4)that naturally sample large areas of the uppercontinental crust. The surface rock method is com-promised by sampling bias, which becomes in-creasingly critical with age, because erosionremoves the upper continental crust with time,and ultramafic and mafic (magnesium- and iron-rich) rocks may be eroded faster than felsic (silica-and aluminum-rich) rocks (1). The terrigenoussediment method cannot provide robust averageconcentrations of soluble elements such as Mg,which are preferentially dissolved and transportedto the oceans during chemical weathering (5).We compiled geochemical data for Archean

shales (including pelites and graywackes), glacialdiamictites (4), and igneous rocks from 18 Archeancratons (6) to demonstrate that Ni/Co and Cr/Znratios provide relatively tight constraints onthe MgO content in the Archean upper conti-nental crust. We use the term “continental crust”here, although the nature of the crust that wasemergent (i.e., exposed to weathering, and hence

the generation of terrigenous sedimentary depos-its) in the Archean may have been very differentfrom the felsic crust that we know today (7).First-row transition metal ratios Ni/Co and

Cr/Zn show positive correlations withMgO contentin igneous and metaigneous rocks from Archeancratons (Fig. 1) due to the differences in their par-tition coefficients between the crystallizing phases

372 22 JANUARY 2016 • VOL 351 ISSUE 6271 sciencemag.org SCIENCE

1Department of Geology, University of Maryland, CollegePark, MD 20742, USA. 2State Key Laboratory of GeologicalProcesses and Mineral Resources, School of Earth Sciences,China University of Geosciences, Wuhan 430074, China.*Corresponding author. E-mail: tangmyes@gmail.com†Present address: Department of Earth Science, University ofCalifornia, Santa Barbara, CA 93106, USA.

Ni/C

o

5

10

15

20

25

MgO, wt.%

Cr/

Zn

0

10

20

30

40

post-Archean igneous rocks

Archean igneous rocks

0 10 20 30 40

Fig. 1. Igneous Ni/Co-MgO and Cr/Zn-MgO dif-ferentiation trends for Archean and post-Archeanrocks.We averaged every 20 samples to reducescatter. Archean igneous trajectories are based oncompiled igneous and metaigneous rocks fromArchean cratons (6); post-Archean trajectoriesare plotted using compiled data from (32).

RESEARCH | REPORTSon June 3, 2020

http://science.sciencemag.org/

Dow

nloaded from

Direct observation of triplet energy transfer from semiconductor nanocrystalsCédric Mongin, Sofia Garakyaraghi, Natalia Razgoniaeva, Mikhail Zamkov and Felix N. Castellano

DOI: 10.1126/science.aad6378 (6271), 369-372.351Science 

, this issue p. 369Sciencecan act as triplet sensitizers.

now show that certain nanoparticleset al.neighbors through energy transfer after absorbing light themselves. Mongin singlet to triplet states is inefficient. Instead, chemists rely on sensitizers, which populate the triplet states of theirengage in a variety of useful reactions but are hard to produce. Quantum mechanics dictates that photo-excitation from

Most molecules adopt a singlet spin configuration: All their electrons are arranged in pairs. Unpaired triplet statesA different way to put triplets in play

ARTICLE TOOLS http://science.sciencemag.org/content/351/6271/369

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2016/01/20/351.6271.369.DC1

REFERENCES

http://science.sciencemag.org/content/351/6271/369#BIBLThis article cites 40 articles, 4 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Copyright © 2016, American Association for the Advancement of Science

on June 3, 2020

http://science.sciencemag.org/

Dow

nloaded from