Figure 1: Three different tehcniques (UV, SAXS and WAXS) coupled on the same experiment following the kinetic of gold NPs in toluene.
We aim at understanding the microscopic mechanisms controlling the formation of nanoparticles. Different types of NPs are studied in the LIONS and among them gold NPs have already been the subject of several achieved PhD (B. Abécassis and F. Hubert) and on-going in-house studies (J. Han) and collaborative ones (S. Gomez from the Univiersity og Vigo). Our general approach is to consider that quantitative kinetic measurements of the size, size distribution and number of NPs with time together with the gold atoms speciation in solution (Au(III), Au(I) and Au-0)) can serve as a basis to develop a better understanding of the mechanisms involved.
The nucleation and growth mechanism of gold nanoparticles are followed at the millisecond time scale using different in situ techniques. Three different systems are presently studied : nanospheres in toluene (NST), nanorods in water (GNR) and nanospheres in water (NSW). The time scale range from one second (NSW) to 600 seconds (GNR). The sample environment are adapted to the time-scale involved. They are (i) a stopped for the most rapid synthesis, (ii) an home-made millifuidic set-up for water rapid systems and (iii) a simple flow through cell for the GNRs.
Three main synchrotron based techniques are used, SAXS WAXS and XANES both with a high time resolution. ESRF (ID02) and SOLEIL (SWING and ODE) have been intensively used to acquire a unique set of data during the nucleation and growth of these objects.
The synthesis was performed in toluene in presence of DDABr helping the solubilization of HAuCl4. Originally, a ligand with an alkyle chain and either a carboxylic or an amine head was added to promote the stabilization of the nanoparticles. Our comprehensive study of these synthese unravelled the more profound role of the ligand. Indeed, the coupling between SAXS and XANES has shown that the size is determined mainly by the rate at which insoluble monomers (Au(0)) are appearing into the solution due to the reduction by the Borohydride. We have shown that the size control empirically observed (the use of decanoic acid as a ligand yields larger particles than decylamine) is in fact due to a direct influence onto the reduction rate. The control of the size is du to a subtle balance between the injection of new insoluble monomers, their consumption by nucleation and their consumption by growth. Taking all the other parameters constant, a general trend is that when the injection (of the same amount of monoeners) is faster, the particles are smaller.
When the injection is slow enough (a few seconds!) the final nanoparticles are larger than 10nm (in diameter) and they spontaneously crystallize in superlattice (of FCC type) in solution.
"Gold superlattice crystallization probed in situ"
Abécassis B. Testard F., Spalla O. Phys. Rev. Lett. (100 (11) 115504 (2008)
"Probing in situ the Nucleation and Growth of Gold Nanoparticles by SAXS"
Abecassis B., Testard F., Spalla, O., Barboux P. Nanoletters 7 (6) 1723-1727 (2007)
Figure 2: Coupling SAXS and XANES. The XANES analysis yields the total amount of gold atoms Au(0) in the system. On the ohter hand, the application of the "Invariant" theorem of the kinetic evolution of the SAXS pattern yiels the amount of gold atoms Au(0) in the nanoparticles versus time. The difference between the two quantities is simply the amount of gold atoms Au(0) transiently free in solution. This amount is definying a supersaturation close to the "burst" of nucleation in nanoparticles.
Figure 3: TEM images of the final distribution of nanorods. A principal component distribution (L, R) is used (each point corresponds to one particle). Two branches exist, one for nanorods and the other for nanospheres. The separation between spheres and rods occurs only a finite size of 5nm.
Two water based systems have been studied, nanorods and nanospheres. For the nanorods, the main question was the occurence of the anisotropy. A quantitative analyse of the the TEM images on final products shows that the longer nanorods are the more anisotropic and that there is no nanorods below a size of 5nm. This is in agreement with a model of differential growth beyond a threshold size. It has also been shown that a complex between CTABr and AgBr strongly adsorb onto the surface of the nanorods and may be at the origin of the difference of growth rate.
Regarding Nanospheres, a neat control of the size can be obtained in the range 5-15nm by playing with the composition of the reducing solution.
“Surfactant (bi)Layers on Gold Nanorods” Gómez-Graña S., Hubert F., Testard F., Guerrero-Martínez A., Grillo I., Liz-Marzán L.M., Spalla O. Langmuir asap (2011)
"Nanorods versus Nanospheres: A Bifurcation Mechanism Revealed by Principal Component TEM Analysis"
Hubert F., Testard F., Rizza G., Spalla O. Langmuir 26 (10) 6887-6891 (2010)
In our general approach the kinetic data (size and size distribution) are quantitatively fitted by a numerical model. This is mandatory to examine whether the microscopic mechanisms supporting the numerical model are realistic. The modelisation is based on the old idea of Lamer. New insoluble monomers are appearing due to the rapid reduction of Au(III) in Au(I) followed by a slower reduction in Au(0). This rate of appearance is measured by the XANES experiment and is the key point. It is not an assumption of the model!! Then, in a second parallel mechanism these monomers associate to form stable nuclei and the Classical Nucleation Theory (CNT) is used to described the rate of nucleation. In a third parallel step, these particles are growing by capturing some of the residual monomers in solution. The growth rate of the nanoparticles depend on the radius due to the Gibbs-Thomson effect. Finally, the conservation of the monomers with time is coupling these three equations. The system of three differential equations is solved with time provided some physical parameters are injected (surface tension, solubility and growth rate contants). The kinetics predictions of the model were positively compared to the experimental measurements.
“Influence of Monomer Feeding on a Fast Gold Nanoparticles Synthesis: Time-Resolved XANES and SAXS Experiments” Abécassis B., Testard F., Kong Q., Baudelet F., Spalla O., Langmuir 26 17 13847-13854 (2010)
Figure 4: Simulation of the nucleation and growth of NPs using three coupled elementary mechanims: (i) appearance of insoluble monomers (ii) nucleation in the framework of the Classical Nucleation Theory (iii) growth of existing particles following depending their radius. The supersaturation in monomers is plotted together with the total number of particles.