Near field scanning optical microscopy

Description Submit all comments and votes

BACKGROUND OF THE INVENTION

The present invention relates to work supported in part by Grant No. ECS 82-00312 of the National Science Foundation.

The present invention relates, in general, to optical microscopy, and more particularly to near-field scanning for high-resolution imaging.

With the advance of submicron technology, the need for a microscope using light for use in the microanalysis of materials has steadily increased. Although various devices, such as electron microscopes, are available for detecting objects with a very high degree of resolution, such prior devices have required that the samples to be observed must be inserted into a vacuum or must be subjected to ionizing radiation. These techniques result in serious damage or destruction of the sample, and, particularly when biological material is being studied, such techniques have been unsatisfactory.

Nondestructive viewing of samples can be obtained with presently available technology, using visible light in two different ranges. At the lower end of the scale, fluorescence spectroscopy, coupled with chemical methods, can be used to determine on a statistical basis the dimensions between objects that are up to about 80 Angstroms apart. At the upper end of the scale, light microscopy, when used in the fluorescence mode, can be used to determine dimensions as small as about one-half the wavelength of the light that is used; that is, down to about 2,500 Angstroms. However, separations between objects, or feature dimensions, of between 80 Angstroms and 2,500 Angstroms are inaccessible when visible wavelengths are used. The ability to determine such dimensions using light microscopy would be very important since, unlike electron microscopy, samples could be studied in their natural environment without resorting to high-vacuum conditions and without the risk of damage. Such actuability would be particularly useful in biological applications where clinical testing or chemical mapping are to be done.

As discussed in copending application Ser. No. 520,041, now U.S. Pat. No. 4,662,747, filed Aug. 3, 1983, and assigned to the assignee of the present application, visible radiation can be transmitted in useful amounts through submicron apertures which are on the order of one-sixteenth of the wavelength of the incident radiation and the radiation emanating through the aperture will be the geometric projection of that aperture. This feature is essentially independent of the wavelength of the incident light. Further, when an aperture is very close to or in contact with an object which is to be imaged, radiation from the object passing through the aperture is the geometric projection of that part of the object which falls within the projection of the aperture. The radiation pattern produced by light passing through a submicron aperture becomes more diffuse as a result of the changing angular distribution of the radiation, which occurs in the Fresnel region. Eventually, a distance is reached where the angular distribution of the radiation pattern becomes constant as a function of distance, so that further motion does not change the shape or size of the pattern. This is known as the far field of the light pattern. Between the aperture and the beginning of the Fresnel region, the radiation is collimated and essentially projecting the shape of the aperture. This region is known as the near-field, and extends for a distance from the surface of the material on which the aperture is formed equal to about one-half the diameter of the aperture.

Over a decade ago, the principle of super-resolution microscopy was demonstrated at microwave frequencies by E. A. Ash and G. Nicholls ("Super-Resolution Aperture Scanning Microscope", Nature 237, p.510, 1972). In their experiment, a grating of 0.5 mm periodicity was imaged with an effective resolution of one-sixtieth the wavelength of the incident radiation. However, until the work described in the above-described copending application Ser. No. 520,041, the applicants therein were unaware of any published attempts to extend this technique to the visible region of the spectrum. Not only did the minute physical dimensions of the optical near-field demand aperture fabrication and micropositioning technologies beyond those available at the time of the Ash, et al, publication, it was also not known whether the results of the microwave experiment could be extended to the visible region. U. C. H. Fischer ("Optical Characteristics of 0.1 um Circular Apertures in a Metal Film as Light Sources for Scanning Ultramicroscopy", J. Vacuum Science Technology, B.3, p.386, 1985) discloses results obtained by scanning a subwavelength aperture over a second, larger aperture. However, the results obtained by that device are difficult to interpret, since the opacity of the metal films used therein was not large, so that the apertures were poorly defined. In addition, coherent monochromatic illumination was used at a grazing incidence, so that a series of standing waves may have been generated to produce the reported results.

A near-field imaging system for use in the far infrared is described by G. A. Massey, et al. ("Subwavelengths Resolution Far-Infrared Microscopy", Applied Optics 24, p.1498, 1985). Although this system may find many applications in the detection of heat transport on a microscopic scale, for example, it does not provide resolution capabilities on a submicron scale. Pohl, et al. ("Optical Stethoscopy: Image Recording With Resolution .lambda./20", Applied Physics Letters 44, p.652, 1984), have developed a system for superresolution microscopy, but the sizes and the structure of the apertures used were not characterized. Furthermore, the manufacturing techniques presented in that article present considerable challenges in the attainment of reproducibility.

The foregoing attempts to implement a near-field scanning technique attest to the difficulty of obtaining success. To further demonstrate the technical challenges inherent in this form of microscopy, the transmission of light through a slit of infinite length in a screen of finite thickness was calculated. The results demonstrated that the radiation passing through such an aperture remains collimated to a distance of at least one-half the slit width and that the extent of the near-field increases with the slit width. Further, the calculations indicated that the near-field energy flux exhibited a close-to-exponential decrease in intensity with increasing distance from the screen. These results suggested that rigid stability requirements would be needed in the direction perpendicular to the surface of the object and of the screen in order to obtain reproducible results.

As described in the aforesaid copending application Ser. No. 520,041, an aperture plate incorporating apertures having diameters on the order of 300 Angstroms has been constructed, and it has been demonstrated that visible light can pass through such apertures, independently of the wavelength of the light. Relatively high transmission is obtained, sufficient to obtain detectable amounts of light using an ordinary microscope illuminator lamp as the light source.

Again, as set forth in Ser. No. 520,041, it was found that spectral phenomena, produced by illuminating an object, also exhibit a near-field radiation pattern; that is, spectral phenomena emanating from an object are essentially perpendicular to the surface from which they emanate, within the near-field region of an aperture used to image the surface. This phenomenon, combined with the use of extremely small apertures, permits observation, in the near-field of an object, of a field of view which is limited to the area of the aperture projected on the surface being observed. As long as the surface is within the near-field of the aperture, the spectral phenomena passing through the aperture will be collimated. An image of the object can be formed if the aperture (or an aperture array) is scanned in a raster-like fashion relative to the object. Such a scanning system has a spatial resolution limited by the aperture diameter instead of by the wavelength of incident light or the spectral phenomena emanating from the surface, and thus can have a resolution on the order of one-tenth to one-sixteenth the wavelength of the incident light.

Although the aperture or aperture array of Ser. No. 520,041 works well as described therein, it has been found that some difficulty has been encountered in attempting to obtain precise observations in those cases where the surface of the object is uneven or is shaped in such a way as to prevent the aperture from being positioned so that the object lies in the near-field of the aperture. This condition severely limits the depth of field of the imaging device, and is of particular concern in the study of biological specimens where it may be desirable to view an object having an extremely uneven surface while still avoiding injury to that object.

When viewing objects through extremely small working distances, the positioning of the viewing aperture becomes extremely critical. In such cases it becomes essential that the instrument be isolated from any environmental vibrations, that changes in materials due to thermal drift be prevented or compensated, and that extremely precise positioning of the aperture be available in all three spatial dimensions. Further, since the light transmitted through a submicron aperture is weak, a sensitive detection system is extremely important, and care must be taken to reduce noise.

SUMMARY OF THE INVENTION

It is, therefore, an object of the present invention to provide an improved method and apparatus for carrying out near-field optical microscopy.

It is a further object of the present invention to provide an improved apparatus for near-field microscopy which apparatus is scannable along an irregular surface.

A further object of the present invention is to provide an improved method for constructing an aperture for use in near-field microscopy.

It is a further object of the resent invention to provide apparatus for scanning in high resolution near-field microscopy.

A still further object of the invention is the provision of an improved optical aperture which is capable of scanning an object in the near-field and which is further capable of detecting low-intensity visible light for near-field microscopy.

Another object of the invention is to provide apparatus for mounting and translating an aperture, the apparatus including feedback techniques for positioning the aperture to permit scanning of rough surfaces in the near-field region.

A particular object of the invention is the provision of a high resolution scanning visible light microscope having a resolution comparable to that of a scanning electron microscope, but also having the ability to study live cells and cellular colonies in their living aqueous environment without the need to destroy the biological system by exposing it to ionizing radiation or by putting it in a vacuum.

Briefly, the present invention is directed to an improved aperture probe for placing a scannable aperture in close proximity to the surface of an object to be imaged. The probe is in the form of a highly tapered, metallized glass pipette, with the aperture being formed at the smallest end thereof. The pipette is formed by heating and pulling a glass tube to taper the pipette and then further heating and pulling it until it breaks. By selecting various wall thicknesses of the glass tube, inner tip diameters of from less than 1,000 Angstroms to 5,000 Angstroms, with outer tip diameters of from 5,000 to 7,500 Angstroms have been reproducibly generated, with the thicker-walled tubes producing smaller inner diameters. An aluminum coating is evaporated onto the outer surface of the pipette so formed to increase the opacity of the glass wall. This evaporation results in a uniformly metallized, tapered pipette with a tip aperture smaller than 1,000 Angstroms, and, preferably, about 500 Angstroms.

The taper of the pipette is of particular importance, for when light is passed down the pipette, it is transmitted through the outer glass wall as well as through the central region. The inner diameter of the pipette rapidly tapers from the nominal dimension of the tube body to dimensions of less than an optical wavelength at the tip. Light transmitted through the pipette in the central opening and in the glass wall remains in a propagating mode throughout the length of the pipette since the outer diameter of the glass wall is larger than the cutoff value that would be expected if it were treated as a classical waveguide. Only at the thin metallized region at the tip of the pipette is the cutoff threshold reached, and, accordingly, a large throughput of light is obtained with only a very small region of decay. Upon leaving the aperture, the radiation once again exists in a propagating mode.

The pipette is mounted on a stable platform such as an optical table which isolates it from vibrations which may be present in the building in which it is mounted. The platform preferably is massive and is mounted on vibration-absorbing air mounts to reduce high frequency vertical disturbances. A honeycomb plate resting on pneumatic supports is then placed on top of the optical table to provide isolation in the intermediate frequency range. Mounted on this plate is a housing which surrounds the optical system. Within the housing is a pair of x and y positioning stages which provide coarse positioning for an object or sample to be observed. These stages may be operated by motor-driven threaded actuators for rough positioning of the sample which is to be imaged. Carried on the x and y positioning stages is a piezoelectric high resolution translation stage for very high resolution positioning of the sample. Such high resolution micropositioning stages are well-known, an example of such a stage being described in "Piezo-driven 50 Micrometer Range Stage with Subnanometer Resolution", Frederick E. Scire, et al., Review of Scientific Instruments 49(12), pp.1735-1740, December 1978. A sample holder stage is mounted on the micropositioner stage.

The pipette is positioned with approximately one micron accuracy over the sample holder, and particularly over the section of the sample to be scanned. A positioning stage is provided to located the pipette in the x-y plane. The pipette is also adjustable in the z direction, a stacked piezoelectric element being used for this purpose. Application of a suitable voltage to the stacked piezoelectric element moves the pipette in the z direction toward and away from the surface of the object to be scanned.

A suitable light source is mounted to illuminate the sample, and light detectors are provided in alignment with the upper end of the pipette. The detectors may be in the form of a fiber optic bundle leading to a photomultiplier or other detector circuitry. Incident light is directed onto the sample and is transmitted through, or reflected by the sample, or fluorescence is produced by the sample, in a direction which is perpendicular to the sample surface in the near-field region of the aperture. This light from the sample is detected by the pipette in its near field, and is directed to the detector circuit, which is in the far field of the aperture.

In one embodiment of the invention, light is transmitted down through the pipette and onto the surface of the sample, with returned light in the form of reflection or fluorescence being returned upwardly through the pipette to the detector circuitry. In another embodiment, light is transmitted continuously through the sample, to the pipette.

During the imaging of a surface, the sample (or the pipette) is moved in the x-y plane in discrete steps with respect to the pipette (or the sample), with the steps preferably being about 150 Angstroms in length. Suitable sensors are provided on the pipette probe to detect the distance between the probe tip and the surface of the sample. Feedback signals to the probe positioning stack adjusts the vertical (z-direction) position of the probe during the scan so as to keep the tip in the near-field region.

The pipette aperture probe and positioning apparatus will provide optical microscopy and dynamic measurements of intermolecular interaction in living systems at a scale that bridges the gap between fluorescence microscopy, with its 2,500 Angstroms resolution, and fluorescence spectroscopy, which allows dimension of less than 80 Angstroms to be measured. The availability of the present probe will generate, at 500 Angstroms resolution, chemically selective images of the molecular constituents of living cells. Furthermore, the present device will permit dynamic measurements at 500 Angstroms resolution and will yield fundamental, new insights in biology and medicine.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and additional objects, features, and advantages of the present invention will become apparent to those of skill in the art from a consideration of the following detailed description of preferred embodiments thereof, taken in conjunction with the accompanying drawings in which:

FIG. 1 is a diagrammatic illustration of the near field region of a submicron aperture;

FIG. 2 is a diagrammatic illustration of the method of making a submicron aperture pipette, in accordance with the present invention;

FIG. 3 is an enlarged, partial, cross-sectional view of a pipette in accordance with the present invention;

FIG. 4 is a diagrammatic illustration of a near-field microscope using the submicron aperture pipette of the present invention; and

FIG. 5 is a side elevation view, in partial section, of a near-field scanning microscope system.

DESCRIPTION OF PREFERRED EMBODIMENTS

The fundamental principle underlying the near-field scanning optical microscope of the present invention is illustrated in FIG. 1, where visible light is depicted by arrows 10 as being normally incident upon a conducting screen 12 containing a small aperture 14 having a diameter less than the wavelength of the incident light 10. Because the screen is completely opaque, the radiation emanating through the aperture and into the region beyond the screen is first collimated to the aperture size rather than to the wavelength of the radiation employed. This collimation occurs in the aperture, but continues into the near-field region 16. Eventually, the effect of diffraction is evidenced as a marked divergence in the emanating radiation, indicated by arrows 18, resulting in a radiation pattern that no longer reproduces the geometrical image of the aperture. This occurs in the far field region, generally indicated at 20.

If an object such as a cell membrane is placed within the near-field region relative to aperture 14, the aperture acts as a light source whose size is not determined by the considerations of geometrical optics, but instead is a projected image of the aperture itself. This light source can be scanned over the object, and detected light, either transmitted through the object or reflected therefrom, can be used to generate a high-resolution image of the area on which the light falls. Because the resolution is dependent upon the aperture size rather than upon the wavelength of the incident radiation, it becomes possible to obtain a resolution of 500 Angstroms or better if a sufficiently small aperture is used. This technique can be applied in air, or in aqueous environments, and uses nonionizing, visible radiation, thus allowing functioning biological systems, as well as other objects, to be imaged at high resolution.

Theoretical calculations of the pattern of radiation in the near-field for light transmitted through a slit of infinite length demonstrate that the radiation remains collimated to a distance of at least one-half the slit width, and that its extent increases with slit width. The near-field energy flux calculations showed a close-to-exponential decrease in intensity with increasing distance from the aperture screen. Thus, the radiation emanating from a 500 Angstroms wide slit would remain collimated to an approximate distance of 250 Angstroms. Similar results are obtained with circular apertures. Because the near-field region is so small, it is apparent that an object to be scanned by such an aperture must be positioned very accurately with respect to the aperture. Furthermore, because the intensity in the near-field is so strongly dependent upon distance, the separation between the aperture and the object must be maintained with a high degree of precision. Because of these strict positioning requirements, the use of the present technique in optical imaging devices requires careful attention to the positioning of the object, maintaining its stability, and enhancing the signal to noise ratio for measurements obtained thereby.

In a typical building environment, vibrations exist at frequencies as low as two to four Hz, although most of the vibrational energy is in the five to 30 Hz range. This can translate into displacements of up to one micron, and if this energy were to be transferred without attenuation to the critical components of a near-field scanning optical microscope, and if the aperture and the object were not held together with sufficient rigidity, then their relative displacement could vary by two to three orders of magnitude greater than the precision required to produce meaningful measurements. Accordingly, it is necessary to isolate the optical microscope from both horizontal and vertical vibrations from the building. In addition, such a system must employ a damping mechanism to ensure that any vibrations which are caused by the motion of the object itself in the course of a scan are also reduced or eliminated. Further, the system must be effectively isolated from acoustic vibrations, which fall in the 20 Hz to 20 kHz range.

Although the thermal expansion coefficients of the various components of a near-field scanning optical microscope system will vary widely with composition and size, it can be expected that such components will expand or contract roughly 0.1 to 1.0 microns for every one degree Centigrade increase or decrease in temperature. Thus, the differential rate of expansion of the aperture relative to an object to be imaged presents a serious obstacle toward obtaining the required micropositioning capabilities for superresolution microscopy, unless the instrument is carefully designed to ensure that the thermal expansion of the aperture relative to the object is rather small or that stringent control is maintained over the temperature of the entire apparatus.

A final requirement for accurate micropositioning of the object with respect to the aperture is to translate the aperture and the object accurately with respect to one another in all three spatial dimensions. The translating system should position the aperture with a precision of roughly 20 Angstroms in the z direction (perpendicular to the plane of the object). It is equally necessary to obtain accurate positioning in the x and y directions (in the plane of the object), since the size of the steps taken during scanning the in the x and y directions limits the resolution of such systems. The step size during a scan should be shorter than one-half of the desired resolution, so a system designed for 500 Angstroms resolution must include positioning control of better than 250 Angstroms in the x and y directions.

The light transmitted through a submicron aperture is weak enough to require the use of sensitive detection electronics, and good detection becomes increasingly important when the aperture scans quickly over the object. Weak signals can particularly be expected in biological applications where contrast differences in a specimen can be quite small. In such applications, a variety of nondestructive methods can be used to increase contrast, ranging from a computer enhancement of images to the fluorescent labelling of specimens.

Since rather weak signals are expected, there can be a significant amount of noise. This noise can be reduced by increasing the period of data collection at any given location, but this is done at the expense of scan speed. Noise can also be introduced if there is any variation in the intensity of the light used to illuminate the sample or the aperture. Furthermore, since the aperture screen may not be completely opaque, any light which is transmitted through it, rather than through this aperture, will contribute to noise. There are also sources of noise related to the translation of the object, for variations in the x-y positioning of the object relative to the aperture can cause an apparent distortion in the image or can smear the contrast and resolution information at a selected point. Of even more significance are the large variations in apparent signal strength which can be caused by slight changes in the aperture to object separation. Finally, noise can be introduced when scanning thick, translucent objects (which are thicker than the near-field region of the aperture), because the system will detect light scatter from the diffraction-limited far field regions as well as from the collimated near-field regions.

Although apertures formed in thin, planar membranes of the type illustrated in FIG. 1 could be used in certain imaging applications, as described in the aforesaid copending application Ser. No. 520,041, it has been found that such apertures have a very limited depth of field due to the limited extend of the collimation of radiation in the near-field region. Because of this limitation, apertures in a planar membrane cannot probe recessed regions in rough surfaces. The present invention overcomes this problem through the provision of a highly tapered, metallized glass pipette which carries at its tip a submicron aperture.

The pipette, and a method of making it to produce the required aperture diameter, are illustrated in FIGS. 2 and 3, to which reference is now made. As shown in FIG. 2, the pipette is formed from a glass tube 24 by means of a simple gravity-driven pipette puller. The puller includes a support clamp 26 for securing tube 24 in a vertical position, and a weight 28 secured to the tube below the clamp. Midway between the clamp 26 and weight 28 is an induction heating coil 30 which surrounds the tube and which, upon application of a suitable current, heats the tube to near the melting point. This causes the tube to elongate in the area of the induction coil due to the force applied by the weight 28. By controlling the temperature of the gl