GB2194115A - Microscopy - Google Patents

Abstract

By incorporating a Pockels cell PC in a polarisation microscope to modulate differentially the orthogonally-polarised component beams enhanced sensitivity to surface variations can be obtained. <IMAGE>

Description

SPECIFICATION Microscopy This invention relates to microscopy.

Many available techniques for microscopy depend, for their proper operation, on object related polarisation variations of the incident light. A simple polarisation microscope, and the Nomarski differential interference microscope are examples of such systems. However, the sensitivity of these techniques is inadequate for imaging many samples, where the variation in the polarisation of the light is small. This can be attributed to the fact that the signal in these systems is obtained at DC and, for small variation of the polarisation, the strength of the signal depends on the square of this variation.

Various techniques have been devised to obtain a linear dependence between the output signal, and the sample structure (Laub, L. : Ac heterodyne profilometer , J. Opt. Soc. Am., 973, 62, p.737; See, C.W., Vaez Iravani, M., and Wickramasinghe, H K.

Scanning differential phase contrast optical microscope application to surface studies, Appl.

Opt., 1985, 24, pp.2373-2379; Huang, C.C : Optical phase profilometer, Opt Eng., 1984, 23, pp.365-370). The operation of most of these systems depends on the extraction of the samplerelated phase of an AC signal, in a phase sensitive detector, by comparing it with a reference signal. We have devised a new technique for performing polarisation/interference contrast microscopy, which provides a signal depending linearly on the variation of polarisation, and thus results in a substantial improvement in sensitivity.

According to the present invention there is provided apparatus for the microscopic investigation of a surface of an object comprising a source of optical radiation, means for resolving radiation from said source into a pair of orthogonally polarised beams, modulation means for differentially modulating said pair of orthogonally polarised beams, means for directing said beams on to said surface and means for measuring said beams after reflection from said surface.

An embodiment of the invention will now be described by way of example with reference to the accompanying drawings in which: Figure 1 is an explanatory diagram, Figure 2 is a schematic arrangement of an optical microscope, and Figures 3 to 8 are micrographs produced from a microscope in accordance with the invention.

Referring to Figure 1 in which the two directions x and y refer to the pass axes of a polariser, and an analyser, respectively. The incoming light, polarised along x, can be resolved into two orthogonally polarised beams, R1 and R2. After reflection off the sample surface, R1 and R2 can be written as R1 = A exp j(ot + e1) 1 R2 = A exp j(Oot + 62) 2 where 01 and 62 are the average optical paths of the two beams. The output of the detector is proportional to the intensity of the total light (after the analyser).Thus 1out : (R1y - R2y)(Rly - R2y)* 3 = 2-2 cos & where se = e2 - e1 4 In the absence of any sample structure, 6 = 0, and the output is zero. A dark background is thus resulted. If, however, the sample alters the value of 6 and 62 unequally, some light is detected. Equation 3 can then be rewritten as

It is thus clear that for small structure, the output is proportional to the square of ##. This results in a severe reduction in sensitivity of the technique in detecting minute surface variations.

Referring back to Figure 1, let us now assume that it is possible to introduce two unequal sinusoidal object independent phase variations into the components. We shall thus have: R1 = A exp j{#ot + flsinx5t + e1) 6 R2 = A -exp j(w,t + 2sine0st + e2) 7 The output of the photodetector is now given by I a (2 - 2[#S(##) J0(se) - 2 sin(##) J1(60) sin xSt + 2 cos (ski) J2 (66) cos was + . . .. 8 where ## = #2 - #1 9 Hence, we arrive at the important result that under these circumstances the signal at s is proportional to sin(dt3), which, for small value of At3, is linearly proportional to JO Accordingly, if the signal at o, is used to image the sample, a greater sensitivity results.

The essential elements of a system in accordance with a specific embodiment of the invention are shown in Figure 2. The output of a linearly polarised He-Ne laser HN is first passed through a polariser P and an electro-optic modulator (Pockels cell) PC, which is driven sinusoidally at a frequency Ws The light is subsequently focused on to the sample S by means of a microscope objective 0. The reflected light is then directed, through an analyser, towards a photodetector.

The design is thus equivalent to a simple scanning polarisation microscope, with the exception of the Pockels-cell.

The presence of the Pockels cell in the system enables us to affect the two orthogonal components of the incoming light differently, which is a requirement for linear operation. Refer ring to equation 8 it is seen that the relative significance of the signals at different frequencies depends on the values of the Bessel function J,(6B). The value of J(p can be adjusted by altering the applied voltage to the Pockels cell, or phase bias, or the position of the cell with respect to the optical beam Maximum amplitude of the fs signal is obtained when it is arranged to be 1.8rad.

The technique described here can also be successfully employed in the implementation of a linear scanning Nomarski differential interference contrast microscope. The configuration of the system in this case is identical to that of Figure 2, except that the final objective is replaced by a Nomarski objective. In this approach, the optical axis of the Nomarski prism is oriented at 45O to that of the incoming light. The sample is, therefore, illuminated by two adjacent foci. It thus follows that the same equations as those pertaining to the linear polarisation microscope hold here, with the signal output given by equation 9.

Similar to the linear polarisation microscope, the linear Nomarski microscope responds to material birefringence. In addition, it will respond to topography, or refractive index variation on the sample.

It is of particular interest to note that in both these systems, the signals at fs' and 2fs' are proportional to sin(oSf3), and cos(##), respectively. One would thus not only expect a greater sensitivity at f,. but also a contrast reversal between the two signals at fs and 2fs.

Let us now calculate the expected sensitivity of the systems. The photodiode current, Io' is given by 1o = rie P 10 h#O where P is the received optical power, and Q the photodiode quantum efficiency. Thus, for a shot noise limited system, we can write for the signal at c95

where F is the amplifier noise factor, Af the system bandwidth. The minimum detectable JO can now be found by equating the S/N to unity. This gives

Thus, for P = 100at, y= 0.8, Af = 10Hz, and F = 2, we have Groin = 5 x 10-7rad 13 In the case of the linear Nomarski microscope, this corresponds to a differential height between the interrogated points of 10-4A.The electro-optic modulator frequency used in our experiment was 100kHz. This enabled the photodetector to be followed by a sufficiently large load impedance. The shot noise limit is thus attained even at relatively low optical powers.

A linear polarisation microscope was set up with the Pockels cell being driven at 100kHz. It was used to examine magnetic domains. To this end, a magnet (extracted from a 1W loud speaker) was polished to a mirror finish (unetched) and was then examined by the system.

Figures 3 and 4 show the micrographs obtained at f5. and 2f respectively. The various domains are perfectly visible on the micrograph of Figure 3. The variation of the brightness on the picture is the manifestation of the relative strength of each of the domains. The micrograph of Figure 4, on the other hand, shows a much weaker signal, as the signal in this case is dependent on the square of the polarisation variation. It is interesting to note that, as expected from theory, a contrast reversal occurs between the two micrographs.

The microscope objective was next replaced with a Nomarski objective. The system was thus transformed into a differential interference contrast microscope. The technique was used to examine the granular structure of a polished, but unetched, stainless steel sample. The micrograph of Figure 5 shows the surface structure of the steel taken at DC, with the Pockels cell removed. In order to prove the sensitivity superiority of the linear Nomarski operation at f,. the optical power was next attenuated by 20dB, and the same field was again examined. Figure 6 is the resulting micrograph at DC. The reduction in the contrast is apparent. The Pockels cell was next inserted into the system, and the experiment repeated, with the optical power kept at the low level. Micrograph 7, obtained at f, shows a much better contrast, and greater detail. Figure 8 shows the same field obtained using the signal at 2fas. We observe a substantial reduction in contrast as compared with that in Figure 7. The improved contrast at f5 is due to the fact that the signal at this frequency is linearly dependent on the differential phase of the sample structure. It is also particularly interesting to note that, as expected, there exists a contrast reversal between the image obtained at fs and that at 2f0 . All these micrographs were recorded in a 30 kHz bandwidth.

Claims (5)

1. Apparatus for the microscopic investigation of a surface of an object comprising a source of optical radiation, means for resolving radiation from said source into a pair of orthogonally polarised beams, modulation means for differentially modulating said pair of orthogonally polarised beams, means for directing said beams on to said surface and means for measuring said beams after reflection from said surface.

2. Apparatus for the microscopic investigation of a surface of an object as claimed in claim 1 wherein said modulation means comprises a Pockels cell.

3. Apparatus for the microscopic investigation of a surface of an object as claimed in either claim 1 or claim 2 wherein said means for directing said beams on to said surface comprises a Nomarski objective.

4. Apparatus for the microscopic investigation of a surface of an object as claimed in claim 3 wherein optical axis of the Nomarski objective is oriented at 45" to that of the incoming radiation.

5. Apparatus for the microscopic investigation of a surface of an object substantially as herein described with reference to and as shown in the accompanying drawings.