Synthesis and Characterization of Surface-Capped, Size-Quantized CdS Clusters. Chemical Control of Cluster Size

Clusters of CdS in the quantum confinement regime C5O-A diameter are prepared in a rational technique whereby the cluster size and its distribution are controlled by chemical means. Competitive reaction chemistry between CdS core cluster growth and surface capping by thiophenolate leads to clusters whose core is essentially sphalerite CdS but whose reactive surface has been passivated by covalently attached phenyl groups. Adjustment of the sulfide to thiophenol ratio during synthesis takes advantage of the competitive reaction rates of these species with Cd ions to control the eventual cluster size. The clusters remain soluble in several organic solvents but may be isolated as stable powders and subsequently redissolved. The Cd "'NMR data for this series of capped clusters confirm the presence of sphalerite CdS as the cluster core and the increasing percentage of Cd involved in this core as the S/SPh ratio increases. Optical properties demonstrate well-behaved absorption edge and emission band shifts with development of exciton features as the clusters grow. Colloidal semiconductor species are currently under intense investigation as examples of nonmolecular materials that demonstrate the effects of quantum confinement.' The enhanced photoreactivity and photocatalysis as well as the predicted effects on nonlinear optical properties of these species has led to a wide range of synthetic approaches to these materials.* The key to any synthetic investigation of this sort must be the careful control of semiconductor cluster size and, even more important, the control of the size distribution. The relatively amorphous character of small clusters prepared in this way necessitates the use of structural probes that provide information even in the absence of long-range periodicity. While X-ra diffraction can offer information for particles in excess of 25 in crystalline dimensions, smaller particles are X-ray amorphous and larger sizes (>lo0 A) can suffer from significant contributions to the line widths by strain broadening. Quantitative interpretation using X-ray diffraction alone is therefore difficult. Among the various techniques suitable for such problems, solidstate N M R has the distinct advantage of providing element-selective, inherently quantitative information about local coordination environments and symmetries. In addition, NMR chemical shifts are also expected to be sensitive to cluster sizes, since the size quantization effects present in small semiconductor clusters should lead to an increase in the average excitation energy parameter in the paramagnetic term of Ramsey's chemical shift theory. This has been recently verified by Duncan and co-workers in a liquid-state "Se NMR study of colloidal CdSe s o l ~ t i o n . ~ We wish to describe a new synthetic approach to such colloids of CdS based on the competitive growth/termination of CdS species in the presence of thiophenol surface-capping agents. The reports by Steigerwald et aL4 using a micellar approach to benzeneselenol-capped CdSe clusters and by Dance et aLs on the preparation of a molecular fragment of sphalerite CdS where a Cdl& core was capped by 16 SPh groups led us to investigate this approach to the rational control of CdS cluster size by capping of the cluster surface by thiophenol-like species. Systematic, detailed optical and NMR behaviors have been revealed. Our N M R data, in addition, complement previous wide-line N M R studies undertaken on bulk cadmium sulfideb6 and add to the extremely limited database presently available for non-oxide chalcogenide systems. Experimental Section Cadmium acetate, thiophenol, and sodium sulfide (anhydrous) were used as received from Alfa Chemical Co. Chemical analyses using E.I. du Pont de Nemours & Co. 'University of California at Santa Barbara. combustion (C + H) and atomic absorption were performed by Galbraith Co., Knoxville, TN. X-ray powder diffraction measurements were performed on a Scintag automated powder diffractometer using Cu Ka radiation. Solid-state "Cd NMR spectra were obtained at 66.69 MHz on a General Electric GN-300 spectrometer, equipped with a multinuclear MAS-NMR probe from Doty Scientific and an Explorer fast digitizer. Normal single-pulse acquisition was used, with 90' pulses of 7-ps length, at a spinning speed between 3.5 and 5.2 kHz. Recycle delays were typically on the order of 10-15 min, resulting in spectra free from saturation effects. Chemical shifts are reported with respect to a sample of liquid Me2Cd. All synthetic procedures and sample manipulations were carried out in an inert-atmosphere (nitrogen) drybox from Vacuum Atmospheres (<lo ppm oxygen, <10 ppm water). Solvents for synthetic procedures and optical spectroscopies were dried and deoxygenated by standard techniques. The compound (NMe4)4(Cd,,$4SPh,6) was prepared in the glovebox by following the published p r ~ e d u r e . ~ Preparation of Clusters. The preparation of all of the cluster species discussed below was essentially identical-the only difference between preparations being in the relative ratio of sulfide to thiophenol used. All preparations were conducted so that [Cd2+] X 2 = [S"] X 2 + [PhSH]. A typical procedure (S:SPh = 1:2) is as follows. Three stock solutions were prepared in the glovebox: (A) 2.67 g of cadmium acetate in 100 mL of methanol, [Cd] = 0.1 M; (B) 0.8 g of sodium sulfide in 50 mL of water and 50 mL of methanol, [S2-] = 0.1 M; and (C) 2.2 mL of thiophenol in 100 mL of methanol, [PhSH] = 0.2 M. These stock solutions are mixed as follows: 50 mL of B and 50 mL of C are stirred well together, and with continuous stirring, 100 mL of solution A is added. This immediately results in a cloudy yellow solution. The solution is stirred for 15 min, then filtered, and suction dried by drawing dry nitrogen through the filter cake for IO min to leave 3.0 g (1) (a) Efros, AI. L.; Efros, A. L. Sou. Phys.-Semicond. (Engl. Trawl . ) 1982,16,772. (b) Brus, L. E. J . Phys. Chem. 1986, 90, 2555, and references therein. (2) (a) Henglein, A. Top. Curr. Chem. 1988, 143, 113. (b) Nozik, A. J.; Williams, F.; Nenadovic, M. T.; Rajh, T.; Micic, 0. I. J . Phys. Chem. 1985, 89, 397. (c) Ramsden, J. J.; Webbcr, S. E.; Gratzel, M. J . Phys. Chem. 1985, 89, 2740. (d) Tricot, Y.-M.; Fendler, J. H. J . Phys. Chem. 1986, 90, 3369. (e) Wang, Y.; Herron. N. J . Phys. Chem. 1987, 91, 5005. (f) Wang, Y.; Herron, N. J . Phys. Chem. 1987, 91, 257. (g) Herron, N.; Wang, Y.; Eddy, M.; Stucky, G . D.; Cox, D. E.; Bein, T.; Moller, K. J . Am. Chem. SOC. 1989, 111, 530. (h) Fojtik, A.; Weller, H.; koch, U.; Henglein, A. Ber. Bunsenges. Phys. Chem. 1984,88, 969. (3) Thayer, A. M.; Steigerwald, M. L.; Duncan, T. M.; Douglass, D. C. Phys. Rev. Lett. 1988, 60, 2673. (4) Steigerwald, M. L.; Alivasatos, A. P.; Gibson, J. M.; Harris, T. D.; Kortan, R.; Muller, A. J.; Thayer, A. M.; Duncan, T. M.; Douglass, D. C.; Brus, L. E. J . Am. Chem. SOC. 1988, 110, 3046. (5) Dance, 1. G.; Choy, A.; Scudder, M. L. J . Am. Chem. SOC. 1984, 106, 6285. (6) Look, D. C. Phys. Status Solidi 1972, BSO, K97. (7) Nolle, A. Z . Naturforsch., A 1978, 33, 666. (8) Eckert, H.; Yesinowski, J. P. J . Am. Chem. SOC. 1986, 108, 2140. 0002-7863/90/1512-1322$02.50/0