New Paper: Ultraviolet

Have you ever looked at the UV optical extinction of plasmon resonant gold nanoparticles?  I mean really looked?  While you wouldn’t expect any interesting plasmonic effects in the UV, some insight might be gained regarding other molecules in the nanoparticle solution. We have been told by colleagues not to bother because the spectrum is “a real mess down there”, but we took that as a challenge.

We recorded extinction spectra with an Ocean Optics PX-2 xenon flash lamp and USB2000 fiber spectrometer optimized for UV detection down to a wavelength of about 185 nm.  To get as far into the UV as possible, we avoided fibers and optics by directly connecting the source and detector with a printed part, so that the light only goes through the quartz cuvette and sample.  With this setup we could measure the extinction reliably down to 200 nm:

Indeed, the spectrum was a real mess down there.  We found a very strong absorption below 220 nm, which we determined was due to the charge-transfer-to-solvent effect (c.t.t.s.) of the bromide ions.  This obscure effect was studied several decades ago, including by fine theorists such as my dean!  The peak is around 180 nm, so we only saw the edge, but it was strong enough that we could use it to measure bromide (and therefore surfactant) concentration.

Once the bromide c.t.t.s. is accounted for, we were able to show that the remaining extinction was due to the UV optical properties of gold.  We switched to nanospheres so we could compare to an exact theoretical expression.  It matched pretty well (20 nm, 60 nm, 80 nm):

The interesting part was that if you go far enough into the UV, you get past the interband absorption of gold, and it behaves like a metal again!  You can see this in the measured dielectric function of gold, which follows the Drude model well below about 275 nm.  You can also see it in the volume-normalized nanoparticle extinction, which is somewhat independent of size in the interband absorptive spectral region (300 – 500 nm), but very dependent on size in the metallic regions:

This explains why other researchers have found that ratios of extinction at certain wavelengths are valuable for tracking nanosphere and nanorod size.  One wavelength is proportional to the amount of gold and the other is affected by the surface.

Ultraviolet Analysis of Gold Nanorod and Nanosphere Solutions
Journal of Physical Chemistry C 2017, 121 p. 5201-5207.


Leave a Reply