Capturing Corrosion, One Pixel at a Time
It’s not just power that corrupts: Moisture or active reagents also can do the trick, at least in the case of metals. Controlling the corrosion that results from such exposure is of great interest. Researchers have tried to understand the underlying process but have not been able to see with enough detail what is going on.
Now Katherine Cimatu and Steven Baldelli from the University of Houston have demonstrated a technique called sum frequency generation imaging microscopy that provides the missing information. They used it to look at the reaction of cyanide with gold and unexpectedly found that the reaction varies from one spot to the next.
“The spatial component of the reaction is somewhat surprising, since often these reactions are considered to be averaged across the surface,” Baldelli said. “Now that we can resolve these reactions and identify them, our knowledge has improved greatly.”
Metals react with corrosive agents to form oxidation products. The result can be something unwanted, such as weakening of the metal. On the other hand, this process is the basis for etching.
Researchers have looked at the corrosion of gold by cyanide because the two are a model of the general interaction between a metal and a corrosive. Gold oxidizes in the presence of cyanide ions, and the reaction has been studied with both spectroscopy and microscopy. Unfortunately, these techniques cannot provide a complete picture of what is going on because they do not supply in situ spatial or spectral information on the gold/cyanide monolayers as they form.
A chemical map shows the corrosion of gold by cyanide. Using sum frequency generation imaging microscopy, researchers acquired images of a gold film’s initial reaction to a corrosive cyanide solution at various wavelengths. On the left is an image at 2105 cm–1. The image on the right is a map of the intensity at 2105 cm–1 across the sample. Images courtesy of Steven Baldelli, University of Houston.
As described in the June 25, 2008, Journal of the American Chemical Society, the Houston scientists’ custom-built sum frequency generation imaging microscope uses two pulsed laser beams that overlap at the surface of a sample and that generate a third beam through a nonlinear process. The third beam has a frequency that is the sum of the two inputs, and it probes the chemistry at the surface of the material.
Two beams make a third for surface probing. In this sum frequency generation imaging microscope, two incoming beams (left) overlap at a sample, producing a third beam that probes the surface chemistry. The resulting signal is collected by optics and routed to a CCD camera.
The investigators used a picosecond Nd:YAG laser from Ekspla of Vilnius, Lithuania, to produce a 1064-nm fundamental beam. With the same beam, they pumped an optical parametric generator/amplifier, yielding a tunable infrared and a 532-nm beam. They sent the 1064-nm and tunable infrared beams at a fixed angle to each other so that they overlapped at a gold film sample. Using filters, lenses and other optical components to remove unwanted light, the scientists collected the image and detected the intensity value in the infrared using a Princeton Instruments 1024 × 1024-pixel CCD camera.
The lateral resolution of the microscope was 10 to 15 μm, with images collected over 1 mm2. Baldelli noted that the microscope’s performance could be improved with a component change. “Better CCD cameras would help with sensitivity and some on the spatial resolution,” he said.
The device collected 41 images in a scan covering 2050 to 2250 cm–1, yielding an intensity value every 5 cm–1. The researchers selected the scan range because there were several peaks within it related to carbon and nitrogen vibrational modes associated with complexes of gold and cyanide.
For their study, the investigators deposited a 100-nm-thick gold film onto a silicon wafer, cut the wafer into centimeter squares and exposed them to a cyanide solution for varying amounts of time. They used sum frequency imaging microscopy on the gold, capturing images at the beginning and after 8 h had passed. They paid particular attention to the values at 2105, 2140, 2170 and 2225 cm–1. The first indicated the presence of cyanide ions linearly bound to the gold, while the others resulted from higher-order complexes.
The beginning set of images showed that the initial reaction created a linear cyanide-gold species. The 8-h images revealed that higher-coordinated gold-cyanide complexes had formed, with up to four cyanide ions bound to a single gold atom. This formation was uneven across the surface, as evidenced by the relative intensities of various peaks in different spots on the gold film. This indicated that the corrosion was surprisingly localized.
The group has continued its work and has studied many monolayer systems via sum frequency generation imaging. The technique has attracted commercial interest, with Baldelli noting that Ekspla is planning to provide a sum frequency generation imaging microscope to its customers. He added that the group’s current device, which was developed with the company, has 2-μm resolution, a significant improvement over the original.
Tools such as this could prove useful in determining what does — and does not — prevent corrosion. “By adding inhibitors to the surface, you could correlate where the corrosion is occurring to the presence, or lack, of the inhibitor,” Baldelli said.
Contact: Steven Baldelli, University of Houston; e-mail: firstname.lastname@example.org.
MORE FROM PHOTONICS MEDIA