Imaging using surface plasmon resonance — electron oscillations induced by light at a metal/dielectric interface — can reveal hidden features such as minute changes in biochemical structures. However, surface plasmon resonance imaging can be fuzzy because the resolution is many times the wavelength of the interrogating light.In a test of the resolution of a surface plasmon resonance imaging system, the effect of various metal films and laser wavelengths is seen. A gold film substrate and a 638-nm laser (top) provide an image that is the least clear, whereas a copper film substrate at the same wavelength has better resolution (middle), as does a gold film substrate and a 532-nm laser (bottom). Scale bars = 6 μm. Reprinted with permission of the American Chemical Society.Researchers at Stanford University in California have demonstrated a microscope that solves this problem. It has a high numerical aperture objective that, when combined with shorter surface plasmon propagation lengths, enables near-diffraction-limited resolution and as clear a picture as possible. “It is really limited by the performance of the high-numerical-aperture objective lens,” said chemistry professor Richard N. Zare.Surface plasmons react to biomolecular binding or other events. Thus, imaging with them permits the label-free readout of DNA or protein microarrays. However, current surface plasmon resonance imaging configurations involve prisms, which limits the numerical aperture and magnification of the system. Consequently, the resolution is many microns, much larger than the diffraction limit.Along with Zare, graduate student Bo Huang and postdoctoral researcher Fang Yu designed, built and tested the system, in which light from a laser passes through collimating and polarizing optics mounted on a translation stage and then through a microscope objective. After reflecting off a metal film that is in contact with the sample, the light is collected by a CCD camera. The translation stage enables the laser offset from the optical axis to be adjusted, ensuring the correct incident angle at the metal film.For their setup, the researchers started with a Nikon Inc. inverted microscope. They used a 638-nm diode laser from CrystaLaser of Reno, Nev., and a 532-nm frequency-doubled Nd:YAG laser from Coherent Inc. of Santa Clara, Calif., for light sources, along with a Roper Scientific Inc. camera for a detector.In their demonstration, they constructed four different substrates with thin films of gold or copper on glass. Zare noted that the technique was not restricted to the two metals. “Certainly, other metals, such as aluminum, are also feasible for achieving this effect,” he said.In one substrate, the metal-coated cover glass was unpatterned. The second had a patterned elastomer stamp pressed against the metal, and the third had a pattern of organic molecules printed on the metal. The last had an uneven plastic film in contact with the metal. The investigators demonstrated that they could image small features and perform angle-resolved surface plasmon resonance imaging on a pixel-by-pixel basis, picking up changes in the dielectric’s thickness of about 1 nm.The resolution was so good that the propagation wavelength of the surface plasmons became the limiting factor in one direction. Solutions for this problem include using a shorter-wavelength laser, a different metal or various incident angles.The researchers have applied for a patent and are planning applications such as simultaneous surface plasmon resonance and surface-plasmon-excited fluorescence imaging.Analytical Chemistry, April 1, 2007, pp. 2979-2983.