Furthermore, although we will describe the basic microscopy systems, we will not deal with all the microscopy features, but we will mainly describe the fluorescent microscopy technique, since it is the currently most important technique in the field of Cell Biology, to whose researchers this chapter is mainly aimed.
Assumed that for microscopy is meant a system of direct vision or by means electronic device (CCD camera or other) that allows a high magnification image, we must understand that, historically, the function of this instrument is twofold, from one hand it allows a magnified view of the observed objects, on the other it allows the human eye to perform such observation under resting conditions. Indeed, the image at the microscope is observed as if it is at infinite distance and, hence, it is not require the contraction of the ciliary muscle. Therefore it is possible to perform observation sessions of long duration with a reduced fatigue for the observer.
The classic scheme of a wide-field microscope is shown in Fig. 41
As you can see the basic structure consists of:
- a light source designed to allow an intense, uniform illumination of the observed area;
- a positioning (x,y,z) system of the sample under observation;
- an objective aimed to project a "X" times magnifiedimage of the specimen on a focal plane, located between 160 and 180 mm (depending on the constructor) from the objective aperture;
- an eyepiece that allows the observation of such focal plane and that introduces an additional magnification factor when the image is observed by means the eyepieces.
In this configuration the magnification of the objective is givenby the ratio between the objective focal length in the direction of the sample and the focal length in direction of the conjugate plane.
Thus for a plane conjugate at 180 mm from the objective pupil and a focal length of 3mm toward the sample, the resulting magnification is of 60 times (60X).
If the eyepiece has a further magnification of 10X, the resulting total magnification to the eye will be of 600 times.
The development of increasingly sophisticated microscopy techniques makes necessary to create a structure of the microscope in order to solve two fundamental problems:
- the use of optical components that require parallel and not divergent beams;
- a large free space that should allow the insertion of several components in the optical path without have each time the inconvenience to change the fundamental optical scheme.
The optimal solution comes from the use of a configuration making use of the infinity objectives, i.e. objectives that do not create an image in a conjugate plane, but they need a second lens (called Tube Lens) to obtain the image of the sample in the desired focal plan. Fig.42.
Even in this case it does not exist a standard value for the focal length of the tube lens, but typical values are between 160 and 180mm.
In configurations with objectives conjugated at infinite, the magnification of the objective is referred with respect to the focal length of the tube lens, hence a 60X objective, for a microscope with a 180 mm tube lens, has a focal length of 3mm.
Fig. 42 - Infinite optical systems with a focus at finite distance
Among the basic components of a microscope, we must remember the diaphragm (i.e. a variable circular or polygonal aperture) of field or of aperture.
The former corresponds to a diaphragm inserted in an intermediate focal plane (placed on the optical path of the illuminator, or of the observation optics) and it is used to avoid that very bright areas of the sample prevent a clear vision of the interest area.
The intensity diaphragm is used to decrease the light intensity and may also improve its uniformity.
The resolving power of a microscope is fundamentally related to the properties of the objective
Currently, the optical design and the manufacturing techniques, make available microscope objectives in which the main aberrations (image defects due to geometric distortions, effects of chromatic aberration, curvature of the focal plane, etc. ..) are reduced to a such low levels that they can be considered negligible with respect to the size of the minimum circle of confusion (blur circle) due to the diffraction.
Therefore, it is usual to find the term "diffraction limited" on the "planar" (where it was compensated the curvature of the focal plane) and "achromatic" (where it was compensated the focus shift due to the refractive index variation with respect to the wavelength ) objectives.
Finally, for applications where the observation is made through a slide, the objective is designed to compensate the optical path variation introduced by the slide thickness.
In a diffraction limited objective, the resolution depends on the numerical aperture of the objective.
There exist several definitions of the objective resolution depending on the numerical aperture, with reference to the Rayleigh criterion we define the resolution, i.e. the ability to distinguish two neighboring points, as:
It is important to recall that in microscopy the number reported on the objective, and indicated as NA, is actually equivalent to 2NA, hence for an objective with NA 1.4 (60X 1.4NA is a common value for high quality, oil immersion objectives) at 500 nm (light of emission), the corresponding nominal resolution is about 220 nm (the minimum distance between two separated points), actually such value is the maximum nowadays obtainable by using high quality instruments at optimal conditions.
We must remember that the real resolution is not a given value, but it depends on many parameters, such as illumination features, the observed sample structure, etc. ..
Finally, it is necessary to indicate that such high numerical apertures values (like 1.4) are, typically, achievable by objectives that use a coupling fluidbetween the first objective lens and the sample, in order to optimize the refractive index variation between the several media and, hence, enabling extremely high values of the acceptance angles of the light beam.
The images observedwith the microscope are due to the interaction between the radiation from the illuminator and the density and/or the refractive index variations among the various components of the sample
Such variations can be extremely weak and therefore give rise to images that are not visible with respect to the intensity of the background illumination.
A possible solution to this problem has been provided by the fluorescence microscopy (or epifluorescence).
Through this technique is not longer observed the reflected or diffused light of the illuminator, but directly the light emitted by the specimen, due to a reaction to the stimulus provided by the illuminator.
In practice the sample, or better, parts of it, are marked with special dyes or associated with chimeric fluorescent proteins (GFP, Green Fluorescent Protein). These, when excited by electromagnetic radiation (light stimulus) with enough energy (i.e. appropriate wavelength), absorb the radiation and (for a phenomenon whose analysis is not the subject of this chapter) re-emit radiation at lower energy (longer wavelength – i.e.: lower energy). As an example we may consider a fluorophore which emits in the green when excited with blue light.
The use of appropriate interference filters (see Chapter 7) makes possible to separate the excitation radiation from the radiation emitted by the sample and, hence,to observe bright images on a dark background.
In the field of cell biology the use of fluorescent media that can be used as markers of constituent parts of the cell, in addition to allow the observation of details not observable by direct techniques, it also allows the observations of functional character, for the chemical and physical specificity of the marker with respect to the marked receptor.
Fig. 44 - Scheme of an epifluorescence microscope
In a normal wide-field microscope is required that the illumination is uniform in order to make visible in a similar way, all portions of the observation field.
This condition, while it is optimal for the observation of surfaces, can be extremely negative in the case of thick translucent bodies, in which are observed internal parts that can be masked by refractive index gradients (separation surfaces) of the sections preceding the examined section.Although these sources of signal are placed on an out-of-focus plans, contribute to the background brightness by creating a diffuse light (haze) that makes unintelligible the image of interest for the observer.
In the case of fluorescence microscopy this effect is even more significant, since it significantly reduces the advantages of the observation in dark field.
The confocal microscopy eliminates (or strongly reduces) such problem by using a point illumination technique, in which the focal illumination point and the observation point are coincident.
Fig. 45 - Scheme of the functioning principle of a confocal microscope
Referring to the scheme of a confocal microscopy, it can be observed that either the excitation beam and the emission beam cross two holes (which may also coincide in a single hole - "pinhole" - "spatial filter"). These holes define the size of the illumination point and the observation area of the sample.
This leads to two main features:
- the intensity of the illumination (radiation of excitation in fluorescent microscopy), being function of the distance from the focal plan, decreases with this distance multiplied by the numerical aperture.
- the collected radiation will be concentrated in the observation hole, since it is the conjugate of the focal plane; while for the front or back plans with respect to the focal plane, an enlargement of the image will occur. Thus only a portion of such signal will cross the observation hole.
Let us imagine two particles of infinitesimal size, the first coinciding with the focal plane and the second aligned to the first one but out-of-focus. For a 60X objective with NA 1.4 in "diffraction limited" conditions and with an observation hole of 30 µm (a value slightly greater than the diameter of the minimum circle of confusion), the effect of the focus shift will lead to a reduction of the intensity of 10 times at a distance from the focal plane of 1 µm; of 41 times at 2 µm and 93 times at 3 µm.
The focus shift in the image plane with respect to the focal plane leads to an attenuation of 8 times for a distance from the focal plane of 1 micron and 32 times for 2 microns.
These values multiplied together lead to an attenuation of 80 times for 1μm of out-of-focus and 1300 for 2μm of out-of-focus.
However, it is useful to notice that the beam profile generated by the fluorescent particle is not flat, but it is the result of the convolution of functions such as sen²x/x² (Figure 46) with the function corresponding to the pinhole, and this latter has dimensions such that the spot that it generates in the focal plane, cannot be considered equal to the diffraction limit.
As a result, at least in the vicinity of the focal plane, the removal of out-of-focus emissions is less than the one stated under the conditions previously described: for a out-of-focus distance of 1 µm is reasonable to divide by a factor 2 the calculated attenuation.
Fig. 46 - sen²x/x²
CONFOCAL LASER SCANNING
The system described above provides a confocal scanning in the plane of the observation point. This scan is usually performed by means the use of two mirrors mounted on galvanometer motors.
The system could work with the light generated by a polychromatic source (e.g. an arc lamp), but considered the small size of the pinhole, in order to have sufficient energy to excite the sample, it is necessary to use laser sources.
In a CLS (Confocal Laser Scanning) system, the resolution along the “z” axis and in the x,y plane is function of the pinhole size, this because the sensor that receives the radiation emitted by the sample, integrates the signal over the whole area of the pinhole.
For a pinhole diameter smaller than one quarter of the blur circle (or half of the Rayleigh resolution), we can assume that the effect of widening of the confusion circle, due to the diffraction, is negligible, and therefore the convolution between the excitation spot and the spot of emission leads to an improvement of the xy resolution of about 1.41 (i.e. the square root of 2).
Obviously in order to make this datum true, it is necessary that the signal is sampled at least twice within the minimum separation distance between two contiguous elements (otherwise we can say that the maximum detectable spatial frequency in the x,y plan, has a period equal to the minimum separation distance , and in order to detect this frequency, the minimum sampling frequency required is twice thespatial frequency).
It follows that in a “raster” scanning system (TV type), the number of horizontal lines shall be twice the required resolution, this in order to provide the same resolution along both the axes.
An example may be clarifying: considering the usualobjective 60X NA1.4, we have a resolution for diffraction equal to 220 nm (for λ = 500nm), if the laser scanning is performed by usinga pinhole not larger than 6μm, the resulting resolution will be roughly 156 nm, hence the signal has to be sampled every 78 nm.
For a field of view of 100 x 100 μm we have about 1300 sampling on the single scanning line along the x-axis, hence we should have 1300 scanning lines along y.
It follows that the maximum electrical frequency will be equal to F = 1300 x 1300 x fps (where fps = frames per second).
Values of the pinhole diameter greater than the one indicated above will lead to a progressive and fast worsening of the resolution.
A final remark should be placed on the method of measurement of the resolution, when measurements are carried out by using a target with vertical bars, since this produces an overestimation of the resolution power due to an artificial increment of the signal, which occurs when the single bar is placed at the center of the spot with respect to two neighbor bars that are intercepted by the edges of the excitation-observation spot.
However, in order to ensure that the frequency of the electrical sampling suits the resolution theoretically provided, a correct assessment of the systems should require to perform two measures with the target oriented either vertically and horizontally and with the scan working at the maximum expected frequency.
The scanning of the target oriented at different angles would be useful to evaluate the possible generation of artifacts due to interference between the scanning lines and the bars of the target.
In principle we can say that the current state of the art of the laser scanning systems, if properly implemented and used, offers the highest spatial resolution and the maximum background noise removal, but it sets significant limits to the acquisition speed (due to the limits of the galvanometer scanning and to the noise due to the high frequency electrical band), therefore its use may result limited on in vivo cells. These, in fact, are dynamical systems that require high speed, and, therefore, very powerful excitation sources, a feature in contrast with the need to avoid high phototoxicity due to the intensity of excitation light.
CONFOCAL SYSTEMS WITH ARRAY SENSOR ( SPINNING DISK STRUCTURED LIGHT)
The confocal systems with array sensor (CCD) are different from the laser scanning systems since the "z" axis and of the x,y plane are solved in two distinct ways, moreover they can generally be used both with laser light and white light sources (as Xenon, Metal Halide lamp) or LEDs.
These systems are provided of a mask placed in a focal plane conjugated to the observed object plane, through this mask travels the excitation radiation and, in some configurations, the emission radiation as well.
Fig. 47 - Spinning Disk
The structure of the excitation light beam produced by the mask defines the resolution along the "z" axis, since the illumination is not uniform in space, but, as in the case of laser scanning, has a maximum in the focal plane, and then it decreases with the square of the distance from the focalplane multiplied by the numerical aperture of the objective used.
In the event that even the emitted radiation travels through the mask, this latter will perform a spatial filtering of the emission, so that the radiation coming from the focus plan will completely pass through the holes of the mask, while the radiation of the out-of-focus plans will be rejected.
The resolution in the x,y plane, in this case, is given by the objective resolution with the only limitation due to the size of the CCD pixels. This because in these systems the holes in the spatial filtering mask are always larger than the diffraction spot (at least when they are used with high resolution objectives, as those with NA greater than 1).
Fig. 48 - CARV scheme
We close this chapter by recalling that are currently being developed systems for structured light that, using spatial filtering or interference techniques, allow to obtain a "sub-diffractive" spatial resolution; resolutions better than 80 nm will be available, but at the expense of observation times currently not suitable for “in vivo” experiments.