MOLECULAR WHITE-LIGHT EMITTER

Abstract

The invention concerns a highly efficient molecular white-light emitter. The invention describes amorphous materials that emit a broadband spectrum of light upon irradiation with an infrared laser. Inorganic nanocrystals form the core of the material and are coated with organic ligands on the surface.

Claims

1. A non-crystalline compound, characterized in that the compound comprises a diamandoid core-structure containing at least one atom of at least one chemical element, chosen from the list comprising C, Si, Ge, Sn, Pb, O, S, Se, Te, whereby the atoms forming the diamandoid core-structure are chosen independently from each other and may or may not carry substituents, whereat a compound comprising either Ge or Sn in combination with S does not comprise the substituent p-Styryl.

2. The compound according to claim 1, characterized in that the core-structure is formed by inorganic nanocrystals of the material comprising at least one chemical element, chosen from the list comprising C, Si, Ge, Sn, Pb, O, S, Se, Te, whereat the nanocrystals are coated with organic ligands on the surface.

3. The compound according to claim 1, characterized in that the compound exhibits a molecular structure according to Formula I, ##STR00002## whereat the atoms M are independently from each other chosen from the list comprising C, Si, Ge, Sn, Pb and atoms X are independently from each other chosen from the list comprising O, S, Se, Te, C and the substituents R are independently from each other chosen from the list comprising aromatic monocyclic substituents, for example Phenyl, Benzyl, Styryl; aromatic polycyclic substituents, for example Naphthyl, Anthryl, Phenanthryl; heteroaromatic monocyclic substituents, for example Pyridyl; heteroaromatic polycyclic substituents, for example aza-Naphthyl, aza-Anthryl, aza-Phenanthryl, diaza-Naphthyl, diaza-Anthryl, diaza-Phenanthryl.

4. The compound according to claim 1, characterized in that the substituents are randomly oriented organic ligands.

5. An electronic device comprising the non-crystalline compound according to claim 1.

6. The electronic device according to claim 5, characterized in that the electronic device is directionally emitting white light.

7. The electronic device according to claim 5, characterized in that the electronic device is selected from the group consisting of laser-diodes, semiconducting laser-diodes, waveguides, light emitting diodes (LED), organic light emitting diodes (OLED), light emitting transistors, LED-screens, OLED-screens, and backlight-devices of TFT-displays.

8. The electronic device according to claim 5, characterized in that the compound is integrated into a monolithic device.

9. The electronic device according to claim 5, characterized in that the compound is applied via vacuum deposition on hydrogen-terminated silicon single crystals and/or on GaAs, whereat the compound forms amorphous layers.

10. The electronic device according to claim 5, characterized in that at least one type of substituent, attached to the diamandoid core-structure is used for further chemical modification and/or covalent attachment of the compound to inorganic materials.

11. A polymer film comprising at least one compound according to claim 1.

12. A device characterized in that the polymer film according to claim 11 is located distant from an infrared laser diode emitting laser light with a wavelength between 800 nm and 1100 nm.

13. The device according to claim 12, characterized in that the polymer film is located between two glass slips.

14. The layer comprising at least one compound according to claim 1.

Description

DETAILED EMBODIMENTS OF THE INVENTION

[0051] Synthesis and Characterization Methods

[0052] All manipulations were performed under argon atmosphere. Acetone was dried and freshly distilled prior to use, as well as all other solvents. Distilled water was degassed freshly prior to use. Trichloro(4-vinylphenyl)tin (R.sup.delocSnCl.sub.3, R.sup.deloc=4-(CH.sub.2CH)C.sub.6H.sub.4) (A) was prepared according to the method reported in the state of the art. [(MeSn).sub.4S.sub.6] (2), [(NpSn).sub.4S.sub.6] (3, Np=1-naphthyl), [(StySn).sub.4S.sub.6] (1, Sty=para-styryl), and [(PhSn).sub.4S.sub.6] (4) are also prepared according to methods reported in the state of the art. Na.sub.2S.9H.sub.2O and all other reagents are purchased from Sigma-Aldrich.

[0053] Melting points were measured in a sealed glass tube on a Krss KSP1 N meltingpoint apparatus. Nuclear magnetic resonance (.sup.1H NMR, .sup.13C NMR, .sup.29Si and .sup.119Sn NMR) measurements were carried out using a Bruker DRX 300 MHz and DRX 500 MHz spectrometer at 298 K. The chemical shifts were quoted in ppm relative to the residual protons of deuterated solvent in .sup.1H NMR and .sup.13C NMR. Me.sub.4Sn was used as external standard for .sup.119Sn and .sup.29Si NMR measurements. Infrared (IR) spectra were recorded on a Bruker Tensor 37. ESI-MS measurements were performed on a Thermo Fischer Scientifics LTQ-FT Ultra mass spectrometer. Energy-dispersive X-ray spectroscopy analysis, EDX, was performed using the EDX device Voyager 4.0 of Noran Instruments coupled with the electron microscope CamScan CS 4DV. Data acquisition was performed with an acceleration voltage of 20 kV and 100 s accumulation time. Elemental analysis was performed on an Elementarvario micro apparatus. RFA is done employing a Bruker Tornado M4.

Synthesis and characterization of 1,3,5,7-Tetrakis(4-vinylphenyl)-2,4,6,8,9,10-hexathia-1,3,5,7-tetrastanna-adamantane, [(R.SUP.deloc.Sn).SUB.4.S.SUB.6.] 1

[0054] Na.sub.2S.9H.sub.2O (0.55 g, 2.3 mmol) was dissolved in a mixture of 5 mL water and 5 mL acetone. A solution of trichloro(4-vinylphenyl)tin (A; R.sup.delocSnCl.sub.3, R.sup.deloc=4-(CH.sub.2CH)C.sub.6H.sub.4), 0.50 g, 1.5 mmol) in 3 mL of acetone was added dropwise at a temperature of 268 K. It was stirred at this temperature for 15 minutes. The resulting colorless precipitate was filtered, washed with water, and dried in high vacuum to become a fine homogeneous powder.

[0055] Yield: 0.36 g (0.33 mmol, 88%)); m.p.: not observed (slow decomposition above 573 K); .sup.1H-NMR (300 MHz, DMF-d.sub.7): 7.80-8.22 p.p.m. (m, 8H), 7.46-7.69 (m, 8H, Ph), 6.79 (dd, J=10.9, 17.8 Hz, 4H, CH), 5.92 (d, J=17.8 Hz, 4H, CH.sub.2), 5.30 (d, J=10.9 Hz, 4H, CH.sub.2); .sup.13C-NMR (75 MHz, DMF-d.sub.7): 114.94, 126.36, 135.03, 135.32, 136.92, 138.94; .sup.119Sn-NMR (187 MHz, DMF-d.sub.7): 258; IR: 2849 (w), 1626 (w), 1586 (w), 1547 (w), 1490 (w), 1386 (m), 1295 (w), 1185 (w), 1068 (w), 1024 (w), 1011 (w), 986 (m), 911 (m), 824 (s), 728 (w), 632 (w), 585 (m), 449 (s) cm.sup.1; HRMS (m/z; see FIG. 5: [M+Na].sup.+ calculated for C.sub.32H.sub.28S.sub.6Sn.sub.4Na, 1102.6497; found, 1102.6488; analysis (% calculated, % found for C.sub.32H.sub.28S.sub.6Sn.sub.4): C (35.60, 35.54), H (2.61, 2.23), S (17.81, 17.34); EDX (calculated, found for 5.sub.6Sn.sub.4): S (1.00, 1.00), Sn (0.67, 0.65).

[0056] As the compound is intrinsically amorphous, the suggested molecular geometry was calculated and validated by means of DFT calculations (see below).

Synthesis of 1,3,5,7-tetraphenyl-2,4,6,8,9,10-hexathia-1,3,5,7-tetrasilana-adamantane, [(PhSi).SUB.4.S.SUB.6.]

[0057] Anhydrous sodium sulfide, Na.sub.2S, (0.906 g, 11.6 mmol) was suspended in tetrahydrofurane (18 mL). At a temperature of 0 C. phenyltrichlorosilane, PhSiCl.sub.3, (1.61 g, 7.75 mmol) was added dropwise under stirring. After two hours the reaction was continued at room temperature for 22 hours. The solvent was evaporated and the residue was extracted with toluene (18 mL). The solvent was slowly evaporated and the product was isolated as single crystalline material.

[0058] Yield: 0.38 g (0.62 mmol, 33% single crystalline yield based on PhSiCl.sub.3). .sup.1H NMR (300 MHz, CDCl.sub.3, 25 C.): 7.49-7.93 (m, 20H) ppm. .sup.13C NMR (75 MHz, CDCl.sub.3, 25 C.): 129.1, 132.6, 133.1, 135.2 ppm. .sup.29Si NMR (99 MHz, CDCl.sub.3, 25 C.): 8.5 ppm. IR: 442 (w), 454 (w), 473 (s), 492 (m), 508 (m), 555 (s), 619 (m), 688 (s), 703 (s), 738 (s), 794 (s), 863 (m), 915 (w), 925 (w), 998 (s), 1014 (s), 1087 (s), 1106 (s), 1184 (w), 1258 (s), 1302 (s), 1334 (s), 1426 (m), 1484 (s), 1587 (s), 2962 (m), 3043 (w), 3066 (w) cm.sup.1. Analysis (% calcd, % found for C.sub.24H.sub.20S.sub.6Si.sub.4): C (36.43, 35.91), H (2.55, 2.52). RFA (calcd, found for S.sub.6Si.sub.4): S (1.00, 1.00), Si (0.67, 0.66).

Synthesis of 1,3,5,7-tetraphenyl-2,4,6,8,9,10-hexathia-1,3,5,7-tetragermana-adamantane, [(PhGe).SUB.4.S.SUB.6.]

[0059] Phenyltrichlorogermane, PhGeCl.sub.3, (1.52 g, 5.93 mmol) was added at room temperature to a solution of sodium sulfide nonahydrate, Na.sub.2S.9H 2 O, (2.12 g, 8.90 mmol) in a solvent-mixture of water (10 mL) and acetone (8 mL). The resulting white precipitate was filtered and washed with water.

[0060] Yield: 0.63 g (0.80 mmol, 42% based on PhGeCl.sub.3). .sup.1H NMR (300 MHz, DMF-d.sub.7, 25 C.): 7.41-8.04 (m, 20H) ppm. .sup.13C NMR (75 MHz, DMF-d7, 25 C.): 130.4, 131.7, 133.0, 133.9 ppm. IR: 420 (m), 457 (m), 537 (m), 617 (w), 690 (s), 732 (s), 815 (s), 834 (m); 883 (w), 929 (m), 996 (w), 1024 (w), 1087 (w), 1099 (w), 1159 (w), 1184 (w), 1260 (w), 1305 (w), 1332 (w), 1432 (m), 1438 (w), 3050 (w), 3070 (w) cm.sup.1. Analysis (% calcd, % found for C.sub.24H.sub.20Ge.sub.4S.sub.6): C (36.43, 36.40), H (2.55, 2.64). RFA (calcd, found for Ge.sub.4S.sub.6): Ge (0.67, 0.64), S (1.00, 1.00).

Synthesis of 1,3,5,7-tetraphenyl-2,4,6,8,9,10-hexathia-1,3,5,7-tetragermana-adamantane, [(PhGe).SUB.4.S.SUB.6.] (5)

[0061] Phenyltrichlorogermane, PhGeCl.sub.3, (1.52 g, 5.93 mmol) is added at room temperature to a solution of sodium sulfide nonahydrate, Na.sub.2S.9H.sub.2O, (2.12 g, 8.90 mmol) in a solvent-mixture of water (10 mL) and acetone (8 mL). The resulting white precipitate is filtered and washed with water.

[0062] Yield: 0.63 g (0.80 mmol, 42% based on PhGeCl.sub.3). .sup.1H NMR (300 MHz, DMF-d.sub.7, 25 C.): 7.41-8.04 (m, 20H) ppm. .sup.13C NMR (75 MHz, DMF-d.sup.7, 25 C.): 130.4, 131.7, 133.0, 133.9 ppm. IR: 420 (m), 457 (m), 537 (m), 617 (w), 690 (s), 732 (s), 815 (s), 834 (m); 883 (w), 929 (m), 996 (w), 1024 (w), 1087 (w), 1099 (w), 1159 (w), 1184 (w), 1260 (w), 1305 (w), 1332 (w), 1432 (m), 1438 (w), 3050 (w), 3070 (w) cm.sup.1. Analysis (% calcd, % found for C.sub.24H.sub.20Ge.sub.4S.sub.6): C (36.43, 36.40), H (2.55, 2.64). RFA (calcd, found for Ge.sub.4S.sub.6): Ge (0.67, 0.64), S (1.00, 1.00).

Synthesis of 1,3,5,7-tetraphenyl-2,4,6,8,9,10-hexathia-1,3,5,7-tetrasilana-adamantane, [(PhSi)4S6] (6)

[0063] Anhydrous sodium sulfide, Na.sub.2S, (0.906 g, 11.6 mmol) is suspended in tetrahydrofurane (18 mL). At a temperature of 0 C. phenyltrichlorosilane, PhSiCl.sub.3, (1.61 g, 7.75 mmol) is added dropwise under stirring. After two hours the reaction is continued at room temperature for 22 hours. The solvent is evaporated and the residue was extracted with toluene (18 mL). The solvent is slowly evaporated and the product is isolated as single crystalline material.

[0064] Yield: 0.38 g (0.62 mmol, 33% single crystalline yield based on PhSiCl.sub.3). .sup.1H NMR (300 MHz, CDCl.sub.3, 25 C.): 7.49-7.93 (m, 20H) ppm. .sup.13C NMR (75 MHz, CDCl.sub.3, 25 C.): 129.1, 132.6, 133.1, 135.2 ppm. .sup.29Si NMR (99 MHz, CDCl.sub.3, 25 C.): 8.5 ppm. IR: 442 (w), 454 (w), 473 (s), 492 (m), 508 (m), 555 (s), 619 (m), 688 (s), 703 (s), 738 (s), 794 (s), 863 (m), 915 (w), 925 (w), 998 (s), 1014 (s), 1087 (s), 1106 (s), 1184 (w), 1258 (s), 1302 (s), 1334 (s), 1426 (m), 1484 (s), 1587 (s), 2962 (m), 3043 (w), 3066 (w) cm.sup.1. Analysis (% calcd, % found for C.sub.24H.sub.20S.sub.6Si.sub.4): C (36.43, 35.91), H (2.55, 2.52). RFA (calcd, found for S.sub.6Si.sub.4): S (1.00, 1.00), Si (0.67, 0.66).

[0065] Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC)

[0066] The TGA/DSC measurements were performed simultaneously on a Netzsch STA 400 with a heating rate of 10 K/min in Ar atmosphere in an Al.sub.2O.sub.3 crucible. The compound decomposes slowly above 573 K in an endothermic process (see FIG. 6). The decomposition is also observable optically by the generation of a grey powder. The maximal mass loss is observed at about 603 K. The mass loss is presumably dominated by the abstraction of organic fragments. At 643 K the speed of abstraction of the organic periphery decreases and continues as a slow process.

[0067] The thermal decomposition of the compounds in inert Ar atmosphere is observed at about 220 C. (2), 270 C. (3), 285 C. (1), 250 C. (4), 310 C. (5) and 370 C. (6), as shown in FIG. 16. In all these cases, the decomposition is an endothermic process.

[0068] Methods of Quantum Chemical Investigations

[0069] For density functional theory (DFT) calculations the program system TURBOMOLE Version 7.0, using the RIDFT program with the BP86 functional and grid size m3 was used. Basis sets were of def2-TZVP quality. For Sn atoms effective core potentials (ECP-28) have been used. No symmetry restrictions (i.e., C1 symmetry) during simultaneous optimizations of geometry and electronic structure were hold. The accuracy of the structures was found within the typical error of the method. Electronic excitations were calculated within the program system by using the ESCF program.

[0070] Quantum-Chemical Investigations of Structural Isomers

[0071] According to the obtained sum formula of (R.sup.delocSn).sub.4S.sub.6, two isomers for the organotin sulfide cluster exist. One isomer exhibits a hetero adamantane type scaffold with pseudo T.sub.d symmetry, the other isomer exhibits a double-decker like scaffold with pseudo D.sub.2 symmetry (see FIG. 7). DFT calculations show that the hetero adamantine topology is energetically favored by about 28 kJ/mol (see Table 1, FIG. 13). For this reason the inventors presume that only the hetero adamantane type scaffold is present at room temperature. This is in agreement to reported phenyl substituted organotin sulfide clusters (RSn).sub.4S.sub.6 (R=C.sub.6F.sub.5, 2,4,6-(CH.sub.3).sub.3C.sub.6H.sub.2, 4-(CH.sub.3)C.sub.6H.sub.4). The rotation of one organic ligand by 180 in the axis of the corresponding CSn bond has a minor influence to the total energy of less than 0.1 kJ/mol. Structural data are provided in Table 2, FIG. 14.

[0072] Calculation of the lowest triplet and singlet excitations, gained from time-dependent DFT (TD-DFT) calculations of the most stable isomer, are in perfect agreement to the observed optical absorption behavior. The lowest singlet excitation exhibits 3.15 eV. The lowest triplet excitation exhibits 2.80 eV, being shifted by 0.3 eV towards lower energy due to fact that electron exchange energy is gained in these case. For higher triplet excitations the energies converge beyond 12 eV which is in good agreement with the 13.46 eV ionization energy used in the simulation of the anharmonic potential.

[0073] Linear Optical Absorption Spectroscopy

[0074] The linear optical absorption behavior was examined by means of UV-visible spectroscopy, measured as powder in reflectance mode under ambient conditions on a Varian Cary5000 spectrometer. Sample spectra are shown in FIG. 17.

[0075] X-Ray Powder Diffraction

[0076] In order to confirm the amorphous nature of compounds 1 to 5 as well as the crystallinity of compound 6, all samples are examined by means of X-ray powder diffraction. The diffractograms are shown in FIG. 18. Powder X-ray diffraction patterns are measured on a StadiMP diffractometer by Stoe equipped with a Mythen 1K silicon strip detector and a Cu-K (=1.54056 ) x-ray source. Samples are measured in transmission between two layers of Scotch Tape (3M).

[0077] Single Crystal X-Ray Crystallography of Compound 6

[0078] Data of the X-ray diffraction analyses are collected on a diffractometer equipped with a STOE imaging plate detector system IPDS2 using Mo.sub.K radiation with graphite monochromatization (=0.71073 ) at 100 K. Structure solution is performed by direct methods, full-matrix-least-squares refinement against F.sup.2 using SHELXTL software. Table 4 (FIG. 19) summarizes data collection and refinement details.

[0079] Density Functional Theory (DFT) Calculations

[0080] Methods of the quantum chemical investigation of compounds 1-6: For the DFT calculations, the program system TURBOMOLE Version 6.5 using the RIDFT program with the BP86 functional and grid size m3 is applied. Basis sets are of def2-TZVP quality. For Ge and Sn atoms, effective core potentials (ECP-28) is employed. The simultaneous optimization of geometric and electronic structures are done without symmetry restrictions (C.sub.1 symmetry). The accuracy of the structures is found within the typical error of the method.

[0081] According to the obtained general sum formula of [(RT).sub.4S.sub.6] (T=Si, Ge, Sn) two isomers for the organotetrel sulfide cluster exist. One isomer exhibits a hetero adamantane type scaffold with pseudo T.sub.d symmetry, the other isomer exhibits a double-decker like scaffold with pseudo D.sub.2 symmetry. DFT calculations show that the hetero adamantane topology is energetically favored in every case with various substituents (RH, Me, Ph) by 20.7 to 41.0 kJ. For compounds 1-6, the calculated values are given in Table 5 (FIG. 21). For compound 5 the hetero-adamantane scaffold is energetically favored by 19.2 kJ/mol. For this reason it can be presumedwithout being confined to a certain theorythat only the hetero adamantane type scaffold is present at room temperature.

[0082] Angular-Resolved Spectroscopy

[0083] The angular-resolved spectroscopy was performed using a 980 nm, 200 mW continuous-wave laser module for excitation. The collimated laser was focused onto the sample using a 5 cm focal-length lens (see FIG. 8, part A). The sample was mounted inside a small vacuum chamber and kept in vacuum at room temperature (293 K). Residual transmitted laser light was blocked using a 3 mm Schott KG3 heat-protective filter-glass. The light emitted from the sample was then collected by a 200 m diameter, 0.12 numerical aperture (NA) optical fiber. The fiber was mounted such, that the center of rotation coincides with the incident laser focus on the sample. Chromatic dispersion and detection of the emitted light was done using a compact spectrometer (OceanOptics USB2000) which was carefully calibrated using a traceable tungsten-halogen standard source.

[0084] Time-Resolved Photoluminescence Spectroscopy

[0085] Time-resolved photoluminescence spectroscopy (TRPL) was performed using a confocal microscopy setup using a streak camera for detection as depicted in FIG. 8, part A. For excitation, we used 100-fs pulses from a Ti:Sapphire laser oscillator with an repetition rate of 78 MHz. Extending its intrinsic operating range, the pulsed could be frequency doubled or tripled. The pulses are focused onto the sample using a 0.5 NA reflective microscope objective. The samples are kept in vacuum inside a cryostat. All measurements are performed at room temperature (293 K). The emission from the sample is collimated in reflection geometry using the same objective and is then imaged either onto a charge-couple device (CCD) camera or onto the entrance slit of a Czerny-Turner-type spectrometer. A spatial resolution <5 m was achieved for optical control by the CCD-camera. Chromatic dispersion was performed using a 40 g/mm grating with a blaze angle of 500 nm. Time-resolution was provided by a standard synchronously scanning Hamamatsu streak-camera. This setup provides a time-resolution of 1.5 ps with an overall time-window of 1.5 ns.

[0086] Steady-State White-Light Spectroscopy

[0087] Steady-state white-light emission spectroscopy was performed using the same optical path and sample holder as for TRPL. A 980 nm, 200 mW laser diode module operating in continuous-wave mode was used for excitation instead of the Ti:sapphire laser. A thermoelectrically-cooled Si-CCD camera is used for detection in the VIS/NIR spectral range. For detection in the NIR/IR range a thermoelectrically cooled (Ga,In)As-CCD camera is used. The cameras are mounted to the second output port of the same imaging spectrometer used for the TRPL experiments. However, here a grating with 122 grooves/mm blazed at 500 nm is used.

[0088] Spectral Response Correction

[0089] All setups were corrected across an extended visible spectral range (300-1100 nm) for the spectral response characteristics to derive the CIE values from the measured spectra. Therefore, a standard traceable tungsten-halogen lamp with a fixed temperature of 2000 K is mounted at a position equivalent to the sample. The spectrum of the lamp was then recorded with both the TRPL and steady state luminescence setup, compared to the supplied black-body spectrum of 2000 K and the correction factor derived.

[0090] Influence of Excitation Wavelength

[0091] To check on the influence of the excitation wavelength multiple spectra of the white-light were obtained using the Ti:sapphire laser tuned to different wavelength, ranging from 725 to 1050 nm. For all these spectra the laser was operating in CW-mode, i.e., not mode-locked. The resulting spectra are given in FIG. 9, part A and do not show significant dependence on the excitation wavelength.

[0092] Influence of Sample Temperature

[0093] To exclude thermal emission as a source of the white-light multiple spectra at different sample temperatures were obtained using the steady-state setup with the 980 nm, 200 mW laser diode for excitation. The excitation density was held constant for all measurements. Here, the sample was cooled by liquid nitrogen in the flow cryostat in contrast to all other measurements. The resulting spectra are shown in FIG. 9, part B. No significant change in spectral line shape is observed ruling out a purely thermal process as the latter would yield changes in the spectral shape similar to those observed for different pump-densities.

[0094] Light Microscopy

[0095] Bright field images of cluster powder layers used to generate supercontinuum were obtained using a standard microscope (Askania RMAS) in reflection geometry. The images were recorded using a Panasonic KR222 CCD-camera. For sample preparation, the powder was dispersed on a coverslip and then capped by another. On a longer-range scale of about 100 m the powder layer is very inhomogeneous with empty spaces between fully covered areas (see FIG. 10, part A). On the fully covered areas, however, the grain size was determined to vary from <1 m up to about 5 m (see FIG. 10, part B-D).

[0096] Microscopy photographs of the compounds 4, 5 and 6 are taken with a stereo microscope (Carl Zeiss-STEMI SV 6) equipped with a standard CMOS camera. The compounds 4 and 5 are obtained as amorphous powders (cf. FIG. 22, a/b). Compound 6 produces large crystallites as shown in FIG. 22, c.

[0097] Measurements of the Nonlinear Optical Response

[0098] The nonlinear responses of the samples are measured using the setup depicted in FIG. 23. The powders are kept inside a small vacuum cell with BK7 windows at pressures below 10.sup.3 mbar. The continuous-wave 980 nm light from a laser diode is focused onto the samples using a 0.5 NA reflective microscope objective in a confocal geometry. A heat absorbing filter (Schott KG3) in the recollimated beam is used to block the residual pump laser. The remaining light, white-light or the second-harmonic, is then focused onto the entrance slit of a Czerny-Turnertype spectrometer where the dispersed spectrum is detected using a thermoelectrically cooled back-illuminated deep-depletion silicon charge-coupled device array sensor.

[0099] Conversion Efficiencies

[0100] Both Mie and Rayleigh scattering are very efficient as the samples are very inhomogeneous and the grain size is in the (sub-)m range (see light-microscopic investigations above and FIG. 10). This leads to large uncertainties in determining the conversion efficiency. Hence, determining the actual net pump intensity that is converted into the supercontinuum is particularly challenging. To assess the uncertainties and avoid any systematic errors the inventors determined the conversion efficiency in several independent ways through different measurements to give the best possible estimation of conversion efficiency. In all cases, a continuous wave Ti:sapphire laser tuned to 980 nm was used for excitation. The laser was focused using a 1 diameter lens and the sample was kept between two coverslips.

[0101] As first method, the efficiency was measured only in forward direction: a high-sensitivity thermal power meter (Thorlabs S401C) was placed behind the sample. Benefit of this method is that power density impinging on the detector is relatively high resulting in an accurate estimation of the emitted power. Different filters were placed in front of the detector in order to ensure that only the desired powers are measured, i.e., a 3 mm thick Schott color glass filter (RG850) discriminated the laser intensity and a 900 nm cut off hard-coated short pass filter with more than five orders of magnitude rejection was used to measure the white light emission.

[0102] This yielded a power of 400 mW with only two coverslips without sample in the beam path. Upon insertion of sample in the beam path, this is reduced to 290 mW; however, it was initially mounted out of focus so that the excitation density is below the supercontinuum threshold and no white light emission occurs. This loss of transmitted laser power is due to the efficient scattering of the sample. Powers of 1.6 mW and 7.36 mW were measured for the white light and the residual laser, respectively, once the sample is placed in the focus and the supercontinuum threshold is overcome. This leads to an efficiency of around 0.5%, but changing the excitation spot size by moving the sample into focus also changes the scattering and thus leads to an underestimation of the efficiency.

[0103] To account for the backscattering, the inventors used a 20 cm diameter integrating sphere. While this reduces the errors due to scattering, one can no longer separate the transmitted and scattered contributions and the power density on the detector will be lower resulting in a larger uncertainty of the measured power. This setup yields a laser power of 2.1 mW at the 5 mm diameter output port below supercontinuum threshold. Above threshold, the inventors find powers of 280 W for the white light and 750 W for the residual laser, yielding an efficiency of 20.745.51%.

[0104] To support these values, the inventors repeated the measurements using a calibrated Si-diode based power meter. While this ensures higher accuracy of the power determination due to its larger sensitivity, a proper spectral response needs to be taken into account. The inventors hence used two band pass filters with transmission of 44010 nm or 632.83 nm to measure the power of two different spectral regions. These values were set in proportion to the overall emitted powers by scaling them with the corresponding emitted spectrum. This way, supercontinuum efficiencies of 16.7110.96% and 9.056.02% are found using the 440 nm filter and the 632 nm filter, respectively.

[0105] Thresholds

[0106] The onset of the supercontinuum generation and the destruction limit of the sample are determined in a setup similar to the steady-state white-light spectroscopy setup discussed above. Now, the spectrometer was replaced by a higher throughput system. This is necessary to ensure more than 4 orders of magnitude in detection sensitivity. The driving laser was focused to a spot size of 0.00138 cm.sup.2. The onset of white light generation was observed for 1 mW, whereas the sample degenerated above 70-80 mW. This yields a threshold value of 0.7 W/cm.sup.2 for the onset of supercontinuum generation, and a destruction threshold of 50 W/cm.sup.2.

[0107] Comparison to Thermal Black-Body Emitter

[0108] To further exclude thermal radiation as the white-light source the respective color temperature for every spectrum obtained in the pump-density dependence series was calculated. Then the emitted powers for every color temperature were estimated by integrating the respective spectrum.

[0109] The setup used to measure the spectra does not yield absolute values of emitted power hence only the relative powers can be compared. The normalized powers plotted vs. color temperatures are shown in FIG. 11. For comparison the T.sup.4 dependence according to Stefan-Boltzmann law that would be expected if the source of the white-light was thermal is also plotted.

[0110] Furthermore a thermal process can be excluded to be the source of white-light radiation by the fact that the compound decomposes above 573 K (see FIG. 6). Hence heating the sample up to the range of 2000 K as it would be required to emit the observed spectra would ultimately result in a destruction of the sample.

[0111] Preparation for Immobilization on Semiconductor Surfaces

[0112] A freshly prepared hydrogen-terminated Si (001) surface or an oxide-terminated GaAs (001) surface and 10 mg 1 were placed next to each other in a glass vessel under Ar atmosphere. The vessel was heated to 523 K for 15 minutes.

[0113] Subsequently, the surface was washed with THF, dried in high vacuum and stored in argon atmosphere. For measurements electron transparent foils of the molecular layers on GaAs (001) and Si (001) substrates were prepared in [110] zone axis of the substrates by conventional mechanical polishing followed by 5 KeV argon ion milling from both sides with an angle of incidence of 4 in a Gatan precision ion polishing system. Final polishing was carried out at 1.7 keV to reduce the amount of amorphous material on the samples.

[0114] Electron Microscopy Deposited on a Si (001) Surface

[0115] The high angle annular dark field (HAADF)-Scanning TEM (STEM) measurements were carried out in a double aberration corrected JEOL 2200 FS, operating at 200 kV. A beam convergence semi-angle of 24 mrad was used and the annular detector was set to collect electrons scattered in the range between 73 and 173 mrads. For EDX spectroscopy in the STEM a Bruker Quantax X-Flash detector was used. The clusters form a perfectly amorphous layer on an H-terminated Si substrate as can be seen from FIG. 12. This molecular layer is separated from the substrate by a region of darker contrast due to the low-Z carbon containing side groups. The inventors find excellent agreement between the contrast seen from the scanning TEM image and the molecular dimensions, suggesting self-limited growth. The arrangement of the clusters is further confirmed by EDX line scans.

[0116] White-Light Spectrum Simulation

[0117] Without being confined to a certain theory, the principle idea behind the white-light generation mechanism is the reemission from an accelerated electron in an anharmonic molecular ground-state potential. For the simulation, the electron experiences the simplified anharmonic potential (U) of the form:

U(x)=m.sub.0.sup.2.Math.x.sup.2+ma.Math.x.sup.3(51)

[0118] Here, x denotes the elongation of the electron from the potential minimum, m is the electron mass, .sub.0 the normal mode of the oscillator, and a the strength of nonlinearity. This potential results in a restoring force of the form:

F(x)=m.sub.0.sup.2.Math.xma.Math.x.sup.2(S2)

[0119] To get a rough estimate of the nonlinearity, we apply Miller's rule, i.e., for elongations equal to the size (d) of the oscillator the linear and nonlinear contributions to the restoring force become comparable. Inserting the size in (S2) leads to the following expression for a:

[00001]$\begin{array}{cc}{m}_0^2.Math.d=\mathrm{ma}.Math.d^2.Math.a=\frac{{}_0^2}{d}& \left(\mathrm{S3}\right)\end{array}$

[0120] Using this substitution the inventors get the equation of motion of the electron:

[00002]$\begin{array}{cc}\stackrel{\mathrm{.Math.}}{x}+2.Math..Math..Math.\stackrel{.}{x}+{}_0^2.Math.x+a.Math.x^2=-\frac{e}{m}.Math.E\left(t\right)& \left(\mathrm{S4}\right)\end{array}$

[0121] Here, is the restoring force of the oscillator that is comparable to damping caused by radiative loss and E(t) is the driving external electric field. This differential equation is solved numerically using the odeint algorithm of the SciPy Phython library. Calculations are performed on a time scale up to 100 ps width a step size of 0.1 fs. The electric field was assumed to be sinusoidal with the frequency of the pump-laser used in the experiment, for the used field-strength as well as the other parameters see: Table 3, FIG. 15. After calculating the movement of the electron the inventors take the Fourier transform of this movement to obtain the frequency components that are comprised in the oscillation of the electron. As a first estimate, this frequency component directly converts to the emitted spectrum.

DESCRIPTION OF THE DRAWINGS

[0122] FIG. 1A: Adamantane-like cluster [(R.sup.delocSn).sub.4S.sub.6] (1, R.sup.deloc=4-(CH.sub.2CH) C.sub.6H.sub.4), with tin and sulfur atoms drawn as dark grey and light grey spheres, respectively; carbon (gray) and hydrogen (white) atoms are given as wires.

[0123] FIG. 1B: Photograph of the as-prepared powder ([(R.sup.delocSn).sub.4S.sub.6] (R.sup.deloc=4-(CH.sub.2CH)C.sub.6H.sub.4)).

[0124] FIG. 1C: Photograph of a polymer film containing the cluster sandwiched between two cover glass slips excited by 800-nm laser light in the bright center spot.

[0125] FIG. 1D: Color temperatures given for various excitation fluencies, as indicated by individual gray-scale data points. The characteristic ideal black-body emission for various temperatures is indicated by the solid line; the square indicates the color temperature of standard emitter at T=2856 K.

[0126] FIG. 2A: Highly directional spatial emission pattern of the white-light spectrum (white) and the CW excitation laser at 980 nm (densely filled).The intensity distribution of a perfect Lambertian emitter (sparsely filled) is given for reference.

[0127] FIG. 2B: White-light spectra for a pump wavelength of 980 nm. The pump power is varied from 6 mW (light gray solid line) to 18 mW (black solid line).The normalized curves for black-body radiation (T=5000 K, dashed line; T=2856 K, spaced dots) and a GaN-based white-light LED (narrow dots) are shown for comparison.

[0128] FIG. 2C: Double-logarithmic plot of the white light input-output characteristics. With a guide to the eye proportional to the 4.sup.th power of the pump density.

[0129] FIG. 3A: Normalized linear absorption spectrum (continuous line), spontaneous emission for UV excitation above the fundamental electronic transition energy (dashed line), and white-light emission spectrum (continuous line marked with squares) for the driving infrared laser on a semilogarithmic scale.

[0130] FIG. 3B: The spontaneous emission (continuous line) decays significantly faster than the white light (continuous line marked with squares).

[0131] FIG. 3C: Schematic illustration portraying the white-light emission due to the accelerated motion of an electron (indicated by continuous-lined trajectory) in the anharmonic electronic ground-state potential (solid curve, marked E.sub.0).

[0132] FIG. 3D: Experimental (solid) and calculated (dotted) white-light emission spectra agree excellently; the scattered part of the driving laser (shaded area) is not included in the simulation.

[0133] FIG. 4A: EDX spectra revealing the contributions of Sn and S in the amorphous cluster layer and Ga and As in the crystalline substrate.

[0134] FIG. 4B: The self-assembled monolayer shows long-range homogeneity and lacks any observable structure, and hence is perfectly amorphous, as can be seen from the high-resolution micrograph; a scaled structure model is overlaid on the micrograph to illustrate the size.

[0135] FIG. 4C: Overlaid at the right side of the micrograph: EDX line scans indicating the distribution of the constituents. The length scale for both (B) and (C) is defined by the right vertical axis.

[0136] FIG. 5: Molecular peak [1+Na]+ in the high resolution ESI(+) mass spectrum of [(R.sup.delocSn).sub.4S.sub.6]. (A) Spectrum measured on a fresh solution of single-crystals of 1 in dichloromethane. (B) Simulated spectrum. Both the maximum abundance and the isotope patterns unambiguously confirm the identity of the compound with the given formula.

[0137] FIG. 6: Thermogravimetric analysis (TG, left vertical axis, dashed line) and differential scanning calorimetry (DSC, right vertical axis, continued line) of [(R.sup.delocSn).sub.4S.sub.6]. The material is thermally stable up to ca. 573 K before an endothermic event is observed that is due to the release of organic ligands to form SnS.sub.2.

[0138] FIG. 7: Minimum structures of [(R.sup.delocSn).sub.4S.sub.6] derived from DFT calculations. [0139] (A) Cluster isomer based on a hetero-adamantane-type topology of the [Sn.sub.4S.sub.6] scaffold. [0140] (B) Cluster isomer based on a so-called double-decker-like architecture of the [Sn.sub.4S.sub.6] scaffold. The cluster based on the hetero-adamantane-type scaffold is energetically favored by about 28 kJ/mol over the other isomer (see Table 1, FIG. 13).

[0141] FIG. 8: (A) Setup for steady-state white-light spectroscopy and time-resolved photoluminescence spectroscopy. The setup for angular-resolved measurements is given in the inset. [0142] (B) Emission characteristics for above band gap excitation. The incident laser (horizontally shaded area) induces the PL (black arrows) that is emitted omnidirectional. [0143] (C) Emission characteristics for below band gap CW-irradiation. The incident laser (horizontally shaded area) induces the white-light emission (vertically shaded area) that is directional but flared around the transmitted laser. Residual emitted PL (dashed arrows) is orders of magnitude lower than for above band gap excitation.

[0144] FIG. 9: (A) White-light emission spectra for various excitation wavelengths indicated by the grey arrows (725-1050 nm). The curves are vertically offset and shifted horizontally for clarity. [0145] (B) White-light emission spectra for various sample temperatures ranging from 70 K up to 325 K for CW-excitation at 980 nm.

[0146] FIG. 10: Reflected light microscopy bright-field images of powder [(R.sup.delocSn).sub.4S.sub.6]. (A-D) Images with increasing magnification: the cluster powder is dispersed on a coverslip and capped by another coverslip.

[0147] FIG. 11: Plot of the normalized emitted powers corresponding to the respective color temperature obtained for the different pump-densities. For comparison, the T.sup.4 dependence according to Stefan-Boltzmann law expected for a thermal emitter is shown (shaded line).

[0148] FIG. 12: Compositional and structural characterization of functionalization characteristics on Si. [0149] (A) EDX spectra. [0150] (B) High-resolution micrograph of the clusters on Si-substrate, overlaid by a space-filling image of the cluster molecule. [0151] (C) EDX line scans. The scatter of the S data is caused by the low signal to noise ratio of the S-related EXD signal.

[0152] FIG. 13: Table 1Total energies of the calculated [(RSn).sub.4S.sub.6] clusters. The [Sn.sub.4S.sub.6] scaffold possesses either hetero-adamantane or double-decker type topology (see FIG. 7).

[0153] FIG. 14: Table 2Calculated structural parameters. The given data represent the structural parameters of the minimum geometry of [(R.sup.delocSn).sub.4S.sub.6 that is based on a hetero-adamantane type [Sn.sub.4S.sub.6] topology.

[0154] FIG. 15: Table 3Parameters used to simulate the spectrum shown in FIG. 3.

[0155] FIG. 16: Thermogravimetric analysis (TGA) of compounds [(PhT).sub.4S.sub.6]: 4 (T=Sn), 5 (T=Ge), and 6 (T=Si).

[0156] FIG. 17: Solid state UV/Vis spectra. Top: Spectra of compounds 2, 3 and 4. Bottom: Spectra of compounds 4, 5 and 6.

[0157] FIG. 18: Top: X-ray powder diffractograms of compounds 1, 2, 3, 4 and 5, plotted one above the other in vertical direction for comparison. Bottom: X-ray powder diffractogram of compound 6 in comparison with a powder pattern simulated from the single crystal structure. The slight angular shift between measured and simulated diffractograms can be attributed to the single crystal measurement being carried out at 100 K while the powder diffractograms were measured at 293 K.

[0158] FIG. 19: Table 4X-ray data collection, structure solution and refinement data of 6.

[0159] FIG. 20: Packing of the molecules in 6, shown as 222 supercells with view along [100] (top), [010] (center), and [001] (bottom). H atoms are omitted for clarity, Si, S, and C atoms are located at the corners. Cluster scaffolds with their inorganic cores are drawn as lines, indicating relative positions of neighboring cluster molecules in the different directions. The relative orientations of the phenyl rings do not indicate n stacking interactions.

[0160] FIG. 21: Table 5Subtraction of the total energies of the calculated hetero-adamantane type clusters (AD) from the double-decker type clusters (DD). (Sty=para-styryl, Np=1-naphthyl), indicating the energetic preference of the hetero-adamantane-type scaffold over the alternative one.

[0161] FIG. 22: Microscopy photographs of the compounds 4, 5 (a, b) and 6 (c).

[0162] FIG. 23: Setup for spectrally resolved measurements of the steady-state nonlinear response.

[0163] FIG. 24: Fragment of (left) the crystal structure (wire representation, shown without H atoms) and (right) the molecular structure (ellipsoids drawn at the 70% probability level) of [(PhSi).sub.4S.sub.6] (6) according to X-ray diffraction studies. Selected structural parameters [, deg]: SiS, 2.1184(5)-2.1476(5); CSi, 1.854(1)-1.855(1); SiSSi, 102.86(2)-104.25(2); SSiS, 111.29(2)-113.27(2); CSiS, 103.85(5)-108.50(5).

[0164] FIG. 25: Emission spectra measured for an excitation wavelength of 980 nm. Compounds 2 and 3 exhibit intense SHG. Compounds 1 and 4 exhibit strong white-light emission (scaled by a factor of 3).

[0165] FIG. 26: Spectra measured for an excitation wavelength of 980 nm. The amorphous compounds 4 and 5 (dashed) exhibit strong white-light emission. The crystalline compound 6 exhibits strong SHG.