Continuous-wave lasing in colloidal quantum dot solids enabled by facet-selective epitaxy

Journal name:
Nature
Volume:
544,
Pages:
75–79
Date published:
DOI:
doi:10.1038/nature21424
Received
Accepted
Published online

Colloidal quantum dots (CQDs) feature a low degeneracy of electronic states at the band edges compared with the corresponding bulk material1, as well as a narrow emission linewidth2, 3. Unfortunately for potential laser applications, this degeneracy is incompletely lifted in the valence band, spreading the hole population among several states at room temperature4, 5, 6. This leads to increased optical gain thresholds, demanding high photoexcitation levels to achieve population inversion (more electrons in excited states than in ground states—the condition for optical gain). This, in turn, increases Auger recombination losses7, limiting the gain lifetime to sub-nanoseconds and preventing steady laser action8, 9. State degeneracy also broadens the photoluminescence linewidth at the single-particle level10. Here we demonstrate a way to decrease the band-edge degeneracy and single-dot photoluminescence linewidth in CQDs by means of uniform biaxial strain. We have developed a synthetic strategy that we term facet-selective epitaxy: we first switch off, and then switch on, shell growth on the (0001) facet of wurtzite CdSe cores, producing asymmetric compressive shells that create built-in biaxial strain, while still maintaining excellent surface passivation (preventing defect formation, which otherwise would cause non-radiative recombination losses). Our synthesis spreads the excitonic fine structure uniformly and sufficiently broadly that it prevents valence-band-edge states from being thermally depopulated. We thereby reduce the optical gain threshold and demonstrate continuous-wave lasing from CQD solids, expanding the library of solution-processed materials11, 12 that may be capable of continuous-wave lasing. The individual CQDs exhibit an ultra-narrow single-dot linewidth, and we successfully propagate this into the ensemble of CQDs.

At a glance

Figures

  1. CdSe CQD band-edge states, state filling and quasi-Fermi-level splitting under hydrostatic and biaxial strain.
    Figure 1: CdSe CQD band-edge states, state filling and quasi-Fermi-level splitting under hydrostatic and biaxial strain.

    Red and blue arrows reveal the presence and absence of compressive strain, respectively. Green and gold balls represent Cd and Se atoms, respectively. On the left, the crystal field and shape anisotropy can lift the hole degeneracy, but the splitting remains comparable to the thermal energy, resulting in broadened photoluminescence and low state filling. This splitting is not affected by hydrostatic strain. On the right, biaxial strain causes additional splitting, concentrating the holes into the lowest-energy levels. Narrower photoluminescence and improved quasi-Fermi level splitting can be realized, lowering the gain threshold. EFe and EFh indicate the quasi-Fermi levels of electrons and holes, respectively, and kT denotes the thermal energy.

  2. Growth of asymmetric CQDs using facet-selective epitaxy.
    Figure 2: Growth of asymmetric CQDs using facet-selective epitaxy.

    a, Facet-selective epitaxy growth mechanism. Green, red and yellow balls represent Cd, Se and S atoms, respectively. TOPS binds weakly to the (000 ) and {10 0} facets but does not bind to the (0001) facet (dots indicate Cd dangling bonds). The primary amine can block the CdS shell growth (pink cross) on the CdSe (0001) facet while keeping the other facets growing (green arrows). Both facets can grow simultaneously when strongly binding octanethiol is used as a sulfur source. b, STEM-EDS mapping confirms the asymmetric shell on the core; the inset shows a HAADF-STEM image of asymmetric dots. c, An HRTEM image (left panel) and the corresponding local lattice constant mapping (middle and right panels) indicate biaxial strain.

  3. Optical characterizations of CdSe–CdS core–shell CQDs.
    Figure 3: Optical characterizations of CdSe–CdS core–shell CQDs.

    a, b, Ensemble absorbance and photoluminescence spectra, Lorentzian fits of single-dot photoluminescence spectra, and tight-binding model simulated exciton fine structure of hydrostatically strained (a) and biaxially strained (b) CQDs, respectively. c, d, Optical gain threshold measurements on the two different types of CQDs. The photexcitation energy was selected to be 2.18 eV to achieve comparable exciton populations with given photoexcitation fluence. The instantaneous total absorbance was collected 27 ps after pulsed excitation. ΔA, A0 and OD denote absorbance change after excitation, ground-state absorbance, and optical density, respectively.

  4. Continuous-wave PC-DFB CQD laser.
    Figure 4: Continuous-wave PC-DFB CQD laser.

    a, Schematic of the PC-DFB device used for lasing. b, Normalized integration of emitted signal (shown in Fig. 4d) as a function of peak power. Insets show emission spectra above and below the lasing threshold. c, Photographs showing emission below and above the threshold, respectively. Above the threshold, bright lasing spots are visible. Images were converted to intensities and displayed with a red colour scale. d, Time traces of normalized emission intensity of the PC-DFB laser as a function of time and incident power while excited with continuous-wave laser.

  5. Numerical simulations.
    Extended Data Fig. 1: Numerical simulations.

    a, b, Poisson distribution of single-exciton (X), biexciton (XX) and multiple excitons in CQD ensembles with different average exciton occupations 〈N〉 under femtosecond (a) and continuous-wave (b) photoexcitations. c, Competition between stimulated emission (blue) and absorption (red) at 〈N〉 = 2.0 exciton population for ensembles. Hydrostatically strained (below the threshold) and biaxially strained (above the threshold) quantum dots. Grey indicates absorption at zero state filling, red indicates absorption of populated dots ensemble, and blue indicates stimulated emission. Black lines indicate excitonic energies and oscillator strengths. d, e, Numerical simulation of absorption bleaching and gain versus exciton occupancy in hydrostatically (d) and biaxially (e) strained CQDs for different hole degeneracies, splitting values and linewidths. f, Numerical simulation showing the dependence of excitation power required to maintain the given occupancy in the continuous-wave regime. A 1.4× gain threshold reduction in terms of excitonic occupancy reduces the required power about 1.7× in the continuous-wave regime. Reducing the gain threshold is critical for realizing continuous-wave lasing. More than 80% of the incident power aimed at achieving population inversion is converted to heat owing to Auger recombination losses.

  6. TEM images and exciton decay dynamics of singly and doubly shelled biaxially strained CQDs.
    Extended Data Fig. 2: TEM images and exciton decay dynamics of singly and doubly shelled biaxially strained CQDs.

    a, Bright-field TEM and HRTEM characterizations of singly shelled asymmetric CQDs. TEM images unveil the hemispherical shape of asymmetric CQDs, which directly confirms that only (0001) facet growth has been blocked by oleylamine when TOPS was used as a sulfur precursor. b, Bright-field TEM and HRTEM characterizations of doubly shelled asymmetric CQDs. To get an overall morphology characterization, doubly shelled CQDs were deposited on lacey-carbon TEM grids with holes to observe CQDs with different orientations. A thorough analysis reveals that these CQDs (nanocrystals) show hexagonal disk-like shapes with two flat faces: viewing from the [0002] zone axis, they show hexagonal or triangular shapes; observing along the [12 0] axis, they show rectangular shapes with two curved sides. The fast Fourier transform patterns (insets) were obtained from the boxed areas. c, Single- and multiple-exciton lifetimes of asymmetric CQDs with single and double shells, respectively. After the second uniform shell growth, fast non-radiative trap recombination is dramatically eliminated, resulting in greatly lengthened single-exciton lifetime (left) and increased PLQY. The multi-exciton dynamics (right) were investigated by ultrafast transient absorption spectroscopy. To achieve the same exciton population, only the CdSe cores were photoexcited (2.18 eV, 570 nm, 507 μJ cm−2). Owing to improved passivation, the amplitude of the slower decay (t2), which can be attributed to trion Auger recombination, decreases notably in doubly shelled asymmetric CQDs, indicating that photoionization has been partially suppressed. As a result, more excitons were retained in doubly shelled CQDs over the entire time regime investigated (0–8 ns). (In all figures, normalizations were performed as follows: maximum scaled to 1, minimum not rescaled.)

  7. Band and exciton fine structures simulations.
    Extended Data Fig. 3: Band and exciton fine structures simulations.

    a, Core-shell structures following the strain relaxation procedure. The left panel shows a tetrahedron shape, with the 4 nm CdSe core exactly at the centre, leading to hydrostatic strain. the right panel shows a disk shape, with the 4 nm CdSe off-centre by 2 nm, leading to biaxial strain. The local lattice constant mapping of the simulated biaxially strained core–shell structure is plotted here for comparison with experimental results in Fig. 2c. b, Wavefunctions of the topmost electron (E1) and hole states (H1, H2, H3, H4) in the biaxially strained CQDs. c, Single-particle states of hydrostatically and biaxially strained CdSe-CdSe core–shell quantum dots calculated with the tight-binding method (h is the manifold of four hole states). d, Exciton fine structures from tight-binding atomistic simulations. The lowest dark state originates from the spin-forbidden configuration based on the heavy hole. The lowest dark–bright splitting depends on the electron–hole exchange interaction, which remains on the order of 1 meV in thick-shell dots and is not affected by the splitting between the heavy-hole and light-hole states; that is, it is insensitive to biaxial strain.

  8. Full absorbance spectra and absorption cross-sections of hydrostatically and biaxially strained CQDs.
    Extended Data Fig. 4: Full absorbance spectra and absorption cross-sections of hydrostatically and biaxially strained CQDs.

    a, b, Full-scale spectra. c, d, Spectra with exciton peaks zoomed in. See Methods for absorption cross-section measurement details.

  9. Absorbance spectra, their second derivatives and photoluminescence spectra of CQDs with varying degrees of splitting.
    Extended Data Fig. 5: Absorbance spectra, their second derivatives and photoluminescence spectra of CQDs with varying degrees of splitting.

    ac, Samples named asymmetric CQD1, CQD2 and CQD3; see synthetic details in the Methods. Asymmetric CQD3 is the single-shell CQD mentioned in the main text. Blue and red arrows highlight the peak positions. d, Photoluminescence spectrum of asymmetric CQD2. ei, Absorbance and photoluminescence spectra evolution during first-shell growth; the resulting product is asymmetric CQD3. e, f, During the growth of the first asymmetric shell, the first exciton peak gradually broadens and then splits into two peaks with increasing reaction time, reaching a maximum splitting of 62 meV. At the same time, the band-edge exciton peak continuously redshifts as a result of increased electron wave function delocalization. g, The progressive splitting is more obvious in the second-derivative absorbance spectra. h, i, The photoluminescence linewidths dramatically decrease after shell growth and photoluminescence peaks redshift as the band-edge exciton peaks shift. j, A summary of data from ef.

  10. Single-dot photoluminescence linewidth measurement.
    Extended Data Fig. 6: Single-dot photoluminescence linewidth measurement.

    a, b, Typical single-dot spectra of hydrostatically (a) and biaxially (b) strained CQDs, with FWHMs of 60 meV and 32 meV, respectively. c, 24 hydrostatically strained CQDs show average photoluminescence linewidth of 63 meV and peak position of 1.96 eV, with standard deviations of 11.8% and 0.67%, respectively. d, The average photoluminescence linewidth and peak position of 20 biaxially strained single dots are 36 meV and 1.95 eV, with standard deviations of 7.2% and 0.5%, respectively. The binning time of the single-dot photoluminescence linewidth measurement is 50 ms.

  11. ASE threshold, and modal gain measurements of CQDs films.
    Extended Data Fig. 7: ASE threshold, and modal gain measurements of CQDs films.

    a, b, Spectra of hydrostatically (a) and biaxially (b) strained CQDs, respectively, with increasing photoexciting power, showing ASE peaks rising above the photoluminescence background. c, Emission as a function of photoexciting peak power density and pulse energy for both hydrostatically and biaxially strained CQDs. A 2 mm × 10 μm stripe with 3.49 eV (355 nm) photoexcitation energy and 1 ns pulse duration was used. The ASE thresholds are 36 μJ cm−2 and 26 μJ cm−2 for hydrostatically and biaxially strained CQDs, respectively. d, Variable stripe length measurements for both hydrostatically and biaxially strained CQDs. Measurements were carried out using a photoexciting energy of four times the threshold value, obtained using a 2 mm × 10 μm stripe. The gain values g are 200 cm−1 and 150 cm−1 for hydrostatically and biaxially strained CQDs, respectively. The lower gain value from biaxially strained CQDs can be explained by the fact that fewer emission states are participating in the optical gain. e, f, ASE threshold measurements of CQD films with 250 fs and 3.49 eV (355 nm) photoexcitation. Thresholds of 22 μJ cm−2 and 14 μJ cm−2 were determined from hydrostatically (e) and biaxially (f) strained CQD films used for measuring the 1 ns ASE threshold.

  12. Single- and multiple-exciton lifetimes of hydrostatically and biaxially strained CQDs, respectively.
    Extended Data Fig. 8: Single- and multiple-exciton lifetimes of hydrostatically and biaxially strained CQDs, respectively.

    a, Owing to the similar total shell volume, both types of core–shell CQDs show similar single-exciton lifetimes. b, The multiple exciton decays with core-only photoexcitation (2.18 eV, 570 nm, 507 μJ cm−2) can be fitted as bi-exponential decays (c): in the fast decay regime, hydrostatically strained core–shell CQDs show slightly longer lifetime than the biaxially strained CQDs, but the opposite is true in the slower decay regime. Since multiple exciton decays are dominated by Auger recombination, the decay rates and amplitudes measured from transient absorption are much higher than those of the single-exciton radiative recombination measured using the TCSPC system. d, e, Multiple exciton decays under different photoexcitation intensities. The bleach signals time traces of hydrostatically strained and biaxially strained CQDs show clear dependence upon photoexcitation power, confirming that Auger recombination is the main decay path.

  13. Continuous-wave PC-DFB CQD lasers with biaxially strained and hydrostatically strained CQDs.
    Extended Data Fig. 9: Continuous-wave PC-DFB CQD lasers with biaxially strained and hydrostatically strained CQDs.

    a, Normalized integration of the emitted signal (shown in Extended Data Fig. 9c) as a function of power under optical photoexcitation at 2.81 eV using a continuous-wave laser. b, Spectra at varying pump powers. c, Time traces of normalized emission intensity of the PC-DFB laser as a function of input excitation power with continuous-wave excitation. d, Emission spectra at powers above and below the threshold for 50 ms pulses and 10 Hz repetition rate, constituting a 50% duty cycle; e, Emission spectra at powers above and below the threshold for 75 ms pulses and 10 Hz repetition rate, constituting a 75% duty cycle. These data were collected by applying a delay to the spectrometer acquisition and measuring the spectra of the last ten microseconds of the pulse. f, HeNe laser (633 nm) signal with electrical interference induced by the acousto-optic modulator driver in the photodiode signal with alternating-current coupling (the measured HeNe laser signal is continuous wave with steady output). g, Electrical signal in the photodiode with no input radiation, showing electrical interference. The oscillation (102 Hz) in these traces and in lasing experiments is consistent, and caused by electrical interference, rather than changes in intensity. h, Emission spectra of PC-DFB lasers made from hydrostatically strained CQDs at powers above and below the threshold for 30 μs pulses and 10 Hz repetition rate, i, Transient behaviour of the lasing intensity for 30 μs at 20 kW cm−2.

Tables

  1. Density-functional-theory ligand-binding energies on different CdSe facets
    Extended Data Table 1: Density-functional-theory ligand-binding energies on different CdSe facets

References

  1. Klimov, V. I. et al. Optical gain and stimulated emission in nanocrystal quantum dots. Science 290, 314317 (2000)
  2. Chen, O. et al. Compact high-quality CdSe-CdS core-shell nanocrystals with narrow emission linewidths and suppressed blinking. Nat. Mater. 12, 445451 (2013)
  3. Dai, X. et al. Solution-processed, high-performance light-emitting diodes based on quantum dots. Nature 515, 9699 (2014)
  4. Efros, A. L. et al. Band-edge exciton in quantum dots of semiconductors with a degenerate valence band: dark and bright exciton states. Phys. Rev. B 54, 48434856 (1996)
  5. Korkusinski, M., Voznyy, O. & Hawrylak, P. Fine structure and size dependence of exciton and biexciton optical spectra in CdSe nanocrystals. Phys. Rev. B 82, 245304 (2010)
  6. Franceschetti, A., Fu, H., Wang, L. W. & Zunger, A. Many-body pseudopotential theory of excitons in InP and CdSe quantum dots. Phys. Rev. B 60, 18191829 (1999)
  7. Klimov, V. I. Multicarrier interactions in semiconductor nanocrystals in relation to the phenomena of auger recombination and carrier multiplication. Annu. Rev. Condens. Matter Phys. 5, 285316 (2014)
  8. Guilhabert, B. et al. Nanosecond colloidal quantum dot lasers for sensing. Opt. Express 22, 73087319 (2014)
  9. McLellan, L. J., Guilhabert, B., Laurand, N & Dawson, M. D. CdS(x)Se(1-x)/ZnS semiconductor nanocrystal laser with sub 10kW/cm2 threshold and 40nJ emission output at 600 nm. Opt. Express 24, A146A153(2016)
  10. Cui, J. et al. Evolution of the single-nanocrystal photoluminescence linewidth with size and shell: implications for exciton-phonon coupling and the optimization of spectral linewidths. Nano Lett. 16, 289296 (2016)
  11. Grim, J. Q. et al. Continuous-wave biexciton lasing at room temperature using solution-processed quantum wells. Nat. Nanotechnol. 9, 891895 (2014)
  12. Iyer, V. H., Mahadevu, R. & Pandey, A. Low threshold quantum dot lasers. J. Phys. Chem. Lett. 7, 12441248 (2016)
  13. Bernard, M. G. A. & Duraffourg, G. Laser conditions in semiconductors. Phys. Status Solidi B 1, 699703 (1961)
  14. Klimov, V. I. et al. Single-exciton optical gain in semiconductor nanocrystals. Nature 447, 441446 (2007)
  15. Yablonovitch, E. & Kane, E. O. Reduction of lasing threshold current-density by the lowering of valence band effective mass. J. Lightwave Technol. 4, 504506 (1986)
  16. Choi, C. L., Koski, K. J., Sivasankar, S. & Alivisatos, A. P. Strain-dependent photoluminescence behavior of CdSe/CdS nanocrystals with spherical, linear, and branched topologies. Nano Lett. 9, 35443549 (2009)
  17. Veilleux, V. et al. Strain-induced effects in colloidal quantum dots: lifetime measurements and blinking statistics. Nanotechnology 21, 134024 (2010)
  18. Cassette, E. et al. Colloidal CdSe/CdS dot-in-plate nanocrystals with 2D-polarized emission. ACS Nano 6, 67416750 (2012)
  19. Christodoulou, S. et al. Band structure engineering via piezoelectric fields in strained anisotropic CdSe/CdS nanocrystals. Nat. Commun. 6, 7905 (2015)
  20. Talapin, D. V. et al. Highly emissive colloidal CdSe/CdS heterostructures of mixed dimensionality. Nano Lett. 3, 16771681 (2003)
  21. Christodoulou, S. et al. Synthesis of highly luminescent wurtzite CdSe/CdS giant-shell nanocrystals using a fast continuous injection route. J. Mater. Chem. C 2, 34393447 (2014)
  22. Cui, J. et al. Direct probe of spectral inhomogeneity reveals synthetic tunability of single-nanocrystal spectral linewidths. Nat. Chem. 5, 602606 (2013)
  23. Tessier, M. D., Javaux, C., Maksimovic, I., Loriette, V. & Dubertret, B. Spectroscopy of single CdSe nanoplatelets. ACS Nano 6, 67516758 (2012)
  24. Tessier, M. D. et al. Spectroscopy of colloidal semiconductor core/shell nanoplatelets with high quantum yield. Nano Lett. 13, 33213328 (2013)
  25. Mahler, B., Nadal, B., Bouet, C., Patriarche, G. & Dubertret, B. Core/shell colloidal semiconductor nanoplatelets. J. Am. Chem. Soc. 134, 1859118598 (2012)
  26. Ithurria, S. & Talapin, D. V. Colloidal atomic layer deposition (c-ALD) using self-limiting reactions at nanocrystal surface coupled to phase transfer between polar and nonpolar media. J. Am. Chem. Soc. 134, 1858518590 (2012)
  27. Dang, C. et al. Red, green and blue lasing enabled by single-exciton gain in colloidal quantum dot films. Nat. Nanotechnol. 7, 335339 (2012)
  28. Adachi, M. M. et al. Microsecond-sustained lasing from colloidal quantum dot solids. Nat. Commun. 6, 8694 (2015)
  29. Carbone, L. et al. Synthesis and micrometer-scale assembly of colloidal CdSe/CdS nanorods prepared by a seeded growth approach. Nano Lett. 7, 29422950 (2007)
  30. Yu, W. W., Qu, L. H., Guo, W. Z. & Peng, X. G. Experimental determination of the extinction coefficient of CdTe, CdSe, and CdS nanocrystals. Chem. Mater. 15, 28542860 (2003)
  31. Coropceanu, I. & Bawendi, M. G. Core/shell quantum dot based luminescent solar concentrators with reduced reabsorption and enhanced efficiency. Nano Lett. 14, 40974101 (2014)
  32. Orfield, N. J., McBride, J. R., Keene, J. D., Davis, L. M. & Rosenthal, S. J. Correlation of atomic structure and photoluminescence of the same quantum dot: pinpointing surface and internal defects that inhibit photoluminescence. ACS Nano 9, 831839 (2015)
  33. She, C. X. et al. Low-threshold stimulated emission using colloidal quantum wells. Nano Lett. 14, 27722777 (2014)
  34. Korkusinski, M., Zielinski, M. & Hawrylak, P. Multiexciton complexes in InAs self-assembled quantum dots. J. Appl. Phys. 105, 122406 (2009)
  35. Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558561 (1993)
  36. Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59, 17581775 (1999)
  37. Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 38653868 (1996)
  38. Jancu, J.-M., Scholz, R., Beltram, F. & Bassani, F. Empirical SPDS* tight-binding calculation for cubic semiconductors: general method and material parameters. Phys. Rev. B 57, 64936507 (1998)
  39. Boykin, T. B., Klimeck, G., Bowen, R. C. & Oyafuso, F. Diagonal parameter shifts due to nearest-neighbor displacements in empirical tight-binding theory. Phys. Rev. B 66, 125207 (2002)
  40. Li, Y. H., Gong, X. G. & Wei, S. H. Ab initio all-electron calculation of absolute volume deformation potentials of IV-IV, III-V, and II-VI semiconductors: the chemical trends. Phys. Rev. B 73, 245206 (2006)
  41. VandeVondele, J. et al. QUICKSTEP: fast and accurate density functional calculations using a mixed Gaussian and plane waves approach. Comput. Phys. Commun. 167, 103128 (2005)
  42. Hartwigsen, C., Goedecker, S. & Hutter, J. Relativistic separable dual-space gaussian pseudopotentials from H to Rn. Phys. Rev. B 58, 36413662 (1998)
  43. VandeVondele, J. & Hutter, J. Gaussian basis sets for accurate calculations on molecular systems in gas and condensed phases. J. Chem. Phys. 127, 114105 (2007)
  44. Beecher, A. N. et al. Atomic structures and gram scale synthesis of three tetrahedral quantum dots. J. Am. Chem. Soc. 136, 1064510653 (2014)
  45. Voznyy, O. et al. A charge-orbital balance picture of doping in colloidal quantum dot solids. ACS Nano 6, 84488455 (2012)
  46. Voznyy, O. Mobile surface traps in CdSe nanocrystals with carboxylic acid ligands. J. Phys. Chem. C 115, 1592715932 (2011)
  47. Cramer, C. J. Essentials of Computational Chemistry: Theories and Models Ch. 10 (John Wiley & Sons, 2013)
  48. Park, Y. S., Bae, W. K., Baker, T., Lim, J. & Klimov, V. I. Effect of auger recombination on lasing in heterostructured quantum dots with engineered core/shell interfaces. Nano Lett. 15, 73197328 (2015)
  49. Korkusinski, M., Voznyy, O. & Hawrylak, P. Theory of highly excited semiconductor nanostructures including auger coupling: exciton-biexciton mixing in CdSe nanocrystals. Phys. Rev. B 84, 155327 (2011)

Download references

Author information

  1. Present address: School of Engineering Science, Simon Fraser University, 8888 University Dr Burnaby, British Columbia V5A 1S6, Canada.

    • Michael M. Adachi
  2. These authors contributed equally to this work.

    • Fengjia Fan,
    • Oleksandr Voznyy,
    • Randy P. Sabatini &
    • Kristopher T. Bicanic

Affiliations

  1. Department of Electrical and Computer Engineering, University of Toronto, 10 King’s College Road, Toronto, Ontario M5S 3G4, Canada

    • Fengjia Fan,
    • Oleksandr Voznyy,
    • Randy P. Sabatini,
    • Kristopher T. Bicanic,
    • Michael M. Adachi,
    • Xiyan Li,
    • Ankit Jain,
    • Rafael Quintero-Bermudez,
    • Mayuran Saravanapavanantham,
    • Min Liu,
    • Sjoerd Hoogland &
    • Edward H. Sargent
  2. The Vanderbilt Institute of Nanoscale Science and Engineering, Vanderbilt University, Nashville, Tennessee 37235, USA

    • James R. McBride,
    • Kemar R. Reid &
    • Sandra J. Rosenthal
  3. Chemistry Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA

    • Young-Shin Park &
    • Victor I. Klimov
  4. Center for High Technology Materials, University of New Mexico, Albuquerque, New Mexico 87131, USA

    • Young-Shin Park
  5. Security and Disruptive Technologies, Emerging Technologies Division, National Research Council, Ottawa, Ontario K1A 0R6, Canada

    • Marek Korkusinski
  6. Physics Department, University of Ottawa, Ottawa, Ontario K1A 0R6, Canada

    • Pawel Hawrylak

Contributions

F.F. and O.V. conceived the idea. F.F. developed the CQD synthesis and performed transient absorption measurements. O.V. carried out theoretical simulations. R.P.S. and K.T.B. designed, fabricated, and characterized the continuous-wave PC-DFB laser devices and films. M.M.A. contributed to PC-DFB laser design, device and film characterizations. J.R.M., K.R.R. and S.J.R. collected STEM-EDS elemental mapping and single-dot photoluminescence linewidth data. Y.-S.P. and V.I.K. contributed to single-dot and low-temperature photoluminescence decay measurements, and modelling of band-edge state degeneracies in relation to gain thresholds. X.L. and M.S. assisted in CQD synthesis optimizations. A.J., M.K. and P.H. contributed to theoretical simulations. R.Q.-B. performed HRTEM lattice analysis. M.L. performed SEM imaging. F.F., O.V., R.P.S., K.T.B. and E.H.S. wrote the manuscript. S.H. and E.H.S. supervised the project. All authors discussed the results and assisted in the preparation of the manuscript.

Competing financial interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to:

Reviewer Information Nature thanks T. Krauss and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author details

Extended data figures and tables

Extended Data Figures

  1. Extended Data Figure 1: Numerical simulations. (249 KB)

    a, b, Poisson distribution of single-exciton (X), biexciton (XX) and multiple excitons in CQD ensembles with different average exciton occupations 〈N〉 under femtosecond (a) and continuous-wave (b) photoexcitations. c, Competition between stimulated emission (blue) and absorption (red) at 〈N〉 = 2.0 exciton population for ensembles. Hydrostatically strained (below the threshold) and biaxially strained (above the threshold) quantum dots. Grey indicates absorption at zero state filling, red indicates absorption of populated dots ensemble, and blue indicates stimulated emission. Black lines indicate excitonic energies and oscillator strengths. d, e, Numerical simulation of absorption bleaching and gain versus exciton occupancy in hydrostatically (d) and biaxially (e) strained CQDs for different hole degeneracies, splitting values and linewidths. f, Numerical simulation showing the dependence of excitation power required to maintain the given occupancy in the continuous-wave regime. A 1.4× gain threshold reduction in terms of excitonic occupancy reduces the required power about 1.7× in the continuous-wave regime. Reducing the gain threshold is critical for realizing continuous-wave lasing. More than 80% of the incident power aimed at achieving population inversion is converted to heat owing to Auger recombination losses.

  2. Extended Data Figure 2: TEM images and exciton decay dynamics of singly and doubly shelled biaxially strained CQDs. (330 KB)

    a, Bright-field TEM and HRTEM characterizations of singly shelled asymmetric CQDs. TEM images unveil the hemispherical shape of asymmetric CQDs, which directly confirms that only (0001) facet growth has been blocked by oleylamine when TOPS was used as a sulfur precursor. b, Bright-field TEM and HRTEM characterizations of doubly shelled asymmetric CQDs. To get an overall morphology characterization, doubly shelled CQDs were deposited on lacey-carbon TEM grids with holes to observe CQDs with different orientations. A thorough analysis reveals that these CQDs (nanocrystals) show hexagonal disk-like shapes with two flat faces: viewing from the [0002] zone axis, they show hexagonal or triangular shapes; observing along the [12 0] axis, they show rectangular shapes with two curved sides. The fast Fourier transform patterns (insets) were obtained from the boxed areas. c, Single- and multiple-exciton lifetimes of asymmetric CQDs with single and double shells, respectively. After the second uniform shell growth, fast non-radiative trap recombination is dramatically eliminated, resulting in greatly lengthened single-exciton lifetime (left) and increased PLQY. The multi-exciton dynamics (right) were investigated by ultrafast transient absorption spectroscopy. To achieve the same exciton population, only the CdSe cores were photoexcited (2.18 eV, 570 nm, 507 μJ cm−2). Owing to improved passivation, the amplitude of the slower decay (t2), which can be attributed to trion Auger recombination, decreases notably in doubly shelled asymmetric CQDs, indicating that photoionization has been partially suppressed. As a result, more excitons were retained in doubly shelled CQDs over the entire time regime investigated (0–8 ns). (In all figures, normalizations were performed as follows: maximum scaled to 1, minimum not rescaled.)

  3. Extended Data Figure 3: Band and exciton fine structures simulations. (343 KB)

    a, Core-shell structures following the strain relaxation procedure. The left panel shows a tetrahedron shape, with the 4 nm CdSe core exactly at the centre, leading to hydrostatic strain. the right panel shows a disk shape, with the 4 nm CdSe off-centre by 2 nm, leading to biaxial strain. The local lattice constant mapping of the simulated biaxially strained core–shell structure is plotted here for comparison with experimental results in Fig. 2c. b, Wavefunctions of the topmost electron (E1) and hole states (H1, H2, H3, H4) in the biaxially strained CQDs. c, Single-particle states of hydrostatically and biaxially strained CdSe-CdSe core–shell quantum dots calculated with the tight-binding method (h is the manifold of four hole states). d, Exciton fine structures from tight-binding atomistic simulations. The lowest dark state originates from the spin-forbidden configuration based on the heavy hole. The lowest dark–bright splitting depends on the electron–hole exchange interaction, which remains on the order of 1 meV in thick-shell dots and is not affected by the splitting between the heavy-hole and light-hole states; that is, it is insensitive to biaxial strain.

  4. Extended Data Figure 4: Full absorbance spectra and absorption cross-sections of hydrostatically and biaxially strained CQDs. (177 KB)

    a, b, Full-scale spectra. c, d, Spectra with exciton peaks zoomed in. See Methods for absorption cross-section measurement details.

  5. Extended Data Figure 5: Absorbance spectra, their second derivatives and photoluminescence spectra of CQDs with varying degrees of splitting. (316 KB)

    ac, Samples named asymmetric CQD1, CQD2 and CQD3; see synthetic details in the Methods. Asymmetric CQD3 is the single-shell CQD mentioned in the main text. Blue and red arrows highlight the peak positions. d, Photoluminescence spectrum of asymmetric CQD2. ei, Absorbance and photoluminescence spectra evolution during first-shell growth; the resulting product is asymmetric CQD3. e, f, During the growth of the first asymmetric shell, the first exciton peak gradually broadens and then splits into two peaks with increasing reaction time, reaching a maximum splitting of 62 meV. At the same time, the band-edge exciton peak continuously redshifts as a result of increased electron wave function delocalization. g, The progressive splitting is more obvious in the second-derivative absorbance spectra. h, i, The photoluminescence linewidths dramatically decrease after shell growth and photoluminescence peaks redshift as the band-edge exciton peaks shift. j, A summary of data from ef.

  6. Extended Data Figure 6: Single-dot photoluminescence linewidth measurement. (201 KB)

    a, b, Typical single-dot spectra of hydrostatically (a) and biaxially (b) strained CQDs, with FWHMs of 60 meV and 32 meV, respectively. c, 24 hydrostatically strained CQDs show average photoluminescence linewidth of 63 meV and peak position of 1.96 eV, with standard deviations of 11.8% and 0.67%, respectively. d, The average photoluminescence linewidth and peak position of 20 biaxially strained single dots are 36 meV and 1.95 eV, with standard deviations of 7.2% and 0.5%, respectively. The binning time of the single-dot photoluminescence linewidth measurement is 50 ms.

  7. Extended Data Figure 7: ASE threshold, and modal gain measurements of CQDs films. (235 KB)

    a, b, Spectra of hydrostatically (a) and biaxially (b) strained CQDs, respectively, with increasing photoexciting power, showing ASE peaks rising above the photoluminescence background. c, Emission as a function of photoexciting peak power density and pulse energy for both hydrostatically and biaxially strained CQDs. A 2 mm × 10 μm stripe with 3.49 eV (355 nm) photoexcitation energy and 1 ns pulse duration was used. The ASE thresholds are 36 μJ cm−2 and 26 μJ cm−2 for hydrostatically and biaxially strained CQDs, respectively. d, Variable stripe length measurements for both hydrostatically and biaxially strained CQDs. Measurements were carried out using a photoexciting energy of four times the threshold value, obtained using a 2 mm × 10 μm stripe. The gain values g are 200 cm−1 and 150 cm−1 for hydrostatically and biaxially strained CQDs, respectively. The lower gain value from biaxially strained CQDs can be explained by the fact that fewer emission states are participating in the optical gain. e, f, ASE threshold measurements of CQD films with 250 fs and 3.49 eV (355 nm) photoexcitation. Thresholds of 22 μJ cm−2 and 14 μJ cm−2 were determined from hydrostatically (e) and biaxially (f) strained CQD films used for measuring the 1 ns ASE threshold.

  8. Extended Data Figure 8: Single- and multiple-exciton lifetimes of hydrostatically and biaxially strained CQDs, respectively. (326 KB)

    a, Owing to the similar total shell volume, both types of core–shell CQDs show similar single-exciton lifetimes. b, The multiple exciton decays with core-only photoexcitation (2.18 eV, 570 nm, 507 μJ cm−2) can be fitted as bi-exponential decays (c): in the fast decay regime, hydrostatically strained core–shell CQDs show slightly longer lifetime than the biaxially strained CQDs, but the opposite is true in the slower decay regime. Since multiple exciton decays are dominated by Auger recombination, the decay rates and amplitudes measured from transient absorption are much higher than those of the single-exciton radiative recombination measured using the TCSPC system. d, e, Multiple exciton decays under different photoexcitation intensities. The bleach signals time traces of hydrostatically strained and biaxially strained CQDs show clear dependence upon photoexcitation power, confirming that Auger recombination is the main decay path.

  9. Extended Data Figure 9: Continuous-wave PC-DFB CQD lasers with biaxially strained and hydrostatically strained CQDs. (486 KB)

    a, Normalized integration of the emitted signal (shown in Extended Data Fig. 9c) as a function of power under optical photoexcitation at 2.81 eV using a continuous-wave laser. b, Spectra at varying pump powers. c, Time traces of normalized emission intensity of the PC-DFB laser as a function of input excitation power with continuous-wave excitation. d, Emission spectra at powers above and below the threshold for 50 ms pulses and 10 Hz repetition rate, constituting a 50% duty cycle; e, Emission spectra at powers above and below the threshold for 75 ms pulses and 10 Hz repetition rate, constituting a 75% duty cycle. These data were collected by applying a delay to the spectrometer acquisition and measuring the spectra of the last ten microseconds of the pulse. f, HeNe laser (633 nm) signal with electrical interference induced by the acousto-optic modulator driver in the photodiode signal with alternating-current coupling (the measured HeNe laser signal is continuous wave with steady output). g, Electrical signal in the photodiode with no input radiation, showing electrical interference. The oscillation (102 Hz) in these traces and in lasing experiments is consistent, and caused by electrical interference, rather than changes in intensity. h, Emission spectra of PC-DFB lasers made from hydrostatically strained CQDs at powers above and below the threshold for 30 μs pulses and 10 Hz repetition rate, i, Transient behaviour of the lasing intensity for 30 μs at 20 kW cm−2.

Extended Data Tables

  1. Extended Data Table 1: Density-functional-theory ligand-binding energies on different CdSe facets (72 KB)

Supplementary information

PDF files

  1. Supplementary Information (1.5 MB)

    This file contains Supplementary Figures 1-5, Supplementary Table 1 and additional references.

Excel files

  1. Supplementary Data 1 (326 KB)

    This file contains the Source Data for Supplementary Figure 1

  2. Supplementary Data 2 (778 KB)

    This file contains the Source Data for Supplementary Figure 2

  3. Supplementary Data 3 (171 KB)

    This file contains the Source Data for Supplementary Figure 3

  4. Supplementary Data 4 (17 KB)

    This file contains the Source Data for Supplementary Figure 5

Additional data