Search

The syntheses, properties, and broad utility of noble metal plasmonic nanomaterials are now well-established. To capitalize on this exceptional utility, mitigate its cost, and potentially expand it, non-noble metal plasmonic materials have become a topic of widespread interest. As new plasmonic materials come online, it is important to understand and assess their ability to generate comparable or complementary plasmonic properties to their noble metal counterparts, including as both sensing and photoredox materials. Here, we study plasmon-driven chemistry on degenerately doped copper selenide (Cu 2- x Se) nanoparticles. In particular, we observe plasmon-driven dimerization of 4-nitrobenzenethiol to 4,4'-dimercaptoazobenzene on Cu 2- x Se surfaces with yields comparable to those observed from noble metal nanoparticles. Overall, our results indicate that doped semiconductor nanoparticles are promising for light-driven chemistry technologies.

Plasmonic materials show great potential for selective photocatalysis under relatively mild reaction conditions. However, the catalytic activity of these plasmonic catalysts can also depend upon the support material that stabilizes the catalysts, where the composition of the catalytic support may change the overall photocatalytic efficiency and yield. It is unknown how changes in the support material may change the plasmon-driven photocatalysis, which may be initiated by plasmon-derived hot carriers, localized heating, or enhanced electromagnetic fields. Herein, we probe the effects of catalytic supports on heating in plasmon-driven catalysis by examining various gold nanoparticle oxide systems. We use ultrafast surface-enhanced Raman thermometry to measure the effective temperature, equivalent to the vibrational kinetic energy, of reporter molecules located between plasmonic gold nanostructures and local environments ranging from ligands to mesoporous silica shells to silica shells. Upon photoexcitation, the transient effective temperature, equivalent to the energy deposited into a vibrational mode, of adsorbed molecules on the silica-coated samples increases, and the energy quickly dissipates within 3 ps. However, the baseline effective temperature that arises from the surface-enhanced Raman spectroscopy probing process depends upon the encapsulant, where the energy deposition differs by 200-300 K between the ligand-coated (citrate or CTAB) and the silica-coated samples. Adsorbates surrounded by a silica shell experience significantly higher effective temperatures than the adsorbates surrounded by ligands or solvent, likely because of the differing effective heat capacities of these media. Taken together, this work shows that a silica support impacts the localized heating of molecular adsorbates on the gold surface and may play a role in enhanced plasmonic photocatalysis because of increased thermal contributions.

Plasmonic materials efficiently convert light to various forms of energies for many applications, including photocatalysis, photovoltaics, and photothermal therapies. In particular, plasmonic photocatalysts hold incredible promise for highly selective sunlight-driven catalysis through the generation of highly energetic holes and electrons used to drive chemical reactions. However, plasmons are also known to generate heat, and the partitioning of photoexcitation energy into hot carriers and heat on molecularly relevant time scales is not well understood, yet plays a crucial role in designing and understanding these photocatalysts. Using an ultrafast surface-enhanced Raman thermometry technique, we probe the effective temperature, equivalent to the mode-specific increase of vibrational kinetic energy, of molecules adsorbed to gold nanoparticle aggregates in the most active hot spots on the picosecond time scale of chemical reactivity. This represents the first measurement of vibrational energy deposition for coupled molecular-plasmonic systems on the picosecond time scale of molecular motion. We find that upon plasmon excitation, the adsorbates in the hot spots undergo an initial energy transfer within several picoseconds that changes the effective temperature of the system by less than 100 K, even at peak flux values 10 8 times stronger than focused sunlight. The energy quickly dissipates from the adsorbates into the surroundings in less than 5 ps, even at the highest values of photoexcitation. This surprisingly modest energy transfer of the most active regions of the plasmonic materials on the ultrafast time scale decisively proves that most plasmonic photocatalysis is not primarily thermally driven.

Hot electrons generated through plasmonic excitations in metal nanostructures show great promise for efficiently driving chemical reactions with light. However, the lifetime, yield, and mechanism of action of plasmon-generated hot electrons involved in a given photocatalytic process are not well understood. Here, we develop ultrafast surface-enhanced Raman scattering (SERS) as a direct probe of plasmon-molecule interactions in the plasmon-catalyzed dimerization of 4-nitrobenzenethiol to p,p'-dimercaptoazobenzene. Ultrafast SERS probing of these molecular reporters in plasmonic hot spots reveals transient Fano resonances, which we attribute to near-field coupling of Stokes-shifted photons to hot electron-driven metal photoluminescence. Surprisingly, we find that hot spots that yield more photoluminescence are much more likely to drive the reaction, which indirectly proves that plasmon-generated hot electrons induce the photochemistry. These ultrafast SERS results provide insight into the relative reactivity of different plasmonic hot spot environments and quantify the ultrafast lifetime of hot electrons involved in plasmon-driven chemistry.

Surface-enhanced Raman spectroscopy (SERS) is a promising technique for in vivo bioanalyte detection, but accurate characterization of SERS biosensors can be challenging due to difficulties in differentiating resonance and surface enhancement contributions to the Raman signal. Here, we quantitate the resonance Raman cross-sections for a commonly used near-infrared SERS dye, 3,3'-diethylthiatricarbocyanine (DTTC). It is typically challenging to measure resonance Raman cross-sections for fluorescent dye molecules due to the overwhelming isoenergetic fluorescence signal. To overcome this issue, we used etalon-based femtosecond stimulated Raman spectroscopy, which is intrinsically designed to acquire a stimulated Raman signal without strong fluorescence or interference from signals resulting from other four-wave mixing pathways. Using this technique, we found that the cross-sections for most of the resonantly enhanced modes in DTTC exceed 10(-25) cm(2)/molecule. These cross-sections lead to high signal magnitude SERS signals from even weakly enhancing SERS substrates, as much of what appears to be a SERS signal is actually coming from the intrinsically strong resonance Raman signal. Our work will lead to a more accurate determination of SERS enhancement factors and SERS substrate characterization in the biologically relevant near-infrared region, ultimately leading to a more widespread use of SERS for biosensing and bioimaging applications.