One fundamental challenge in fluorescence microscopy including and confocal microscopy is the separation of excitation light, e.g., from a laser, and fluorescence light from the sample. The excitation light should pass through to the sample unhindered, while none of it must contaminate the fluorescence signal which is weaker by a factor of 10-5 (see Figure 1). For comparison, this ratio is similar to the light intensities of one small candle versus twenty 1000 W halogen lamps. To ensure that fluorescence light can be detected without contamination from excitation light, the beam splitter is the crucial component for image contrast in the beam path. For a general overview of different types of beam splitters in confocal microscopy, please refer to the corresponding article on Leica Science Lab by Rolf Borlinghaus. The following will focus on specific types of beam splitters used in the Leica TCS SP8.
Figure 1: Beam splitter in fluorescence microscopy. Excitation light (blue arrow) is coupled into the light path using a dichroic mirror and passed on to the sample. Part of it is reflected back to the beam splitter where it is blocked. The majority excites fluorescence (red dashed arrow) which passes through the beam splitter to be detected. Note the dramatic difference in intensity between excitation and emission light by five orders of magnitude.
LIAchroics are an advancement of dichroic beam splitters. The latter behave like a semi-transparent mirror which is wavelength selective, e.g., blue light gets reflected, while green light can pass through. This means that dichroic mirrors, as they are also called, are selected according to a fluorescent dye’s Stokes shift. The specific spectral properties of a beam splitter are determined by a thin layer of optical coatings on a glass substrate. Dichroic beam splitters are interference filters. The coating is fixed, hence, for each laser/dye combination a dedicated beam splitter is required. To accomplish this, a set of filters is usually placed on a movable slider or a disc to move into and out of the beam path as needed. As shown in Figure 1, the traditional design uses a 45° angle for the dichroic mirror, which has advantages for geometric separation of excitation and emission light with their beams being orthogonal. However, this design is not equally permissive to different angles of (linearly) polarized light. Consequently, a part of the fluorescence light is lost, the polarization plane of which happens to have the “wrong” orientation. The explanation is the so-called Brewster’s angle. Whether light is transmitted or reflected on an air-glass interface depends on its angle of incidence. For air-glass interfaces Brewster’s angle is around 56° – too close to the 45° typically used for dichroic mirrors. Nice visualizations of air-glass interfaces by Srihari Angaluri and Kiril N Vidimce are found here. When the angle of the incident light is much smaller than 45°, this effect is eliminated. Hence, LIAchroic beam splitters, or low incident light dichroics, offer improved contrast and sensitivity compared to previous beam splitters. LIAchroics are custom-designed and produced in-house to guarantee affordable beam splitting efficiency.
Despite the recent advances in production of dichroic mirrors, two inherent handicaps still remain: Being filters, they always loose some fluorescence light and their confinement to a certain wavelength limits the flexibility of the confocal instrument. Light loss in dichroics occurs not only because of transmission smaller than 100% above the splitting wavelength but mostly due to filter characteristics close to this wavelength. Typically, dichroic beam splitters suppress a band around the laser wavelength of about 20 nm width. Since this spectral region often contains the emission maximum of a dye’s spectrum, light loss will occur (see Figure 2, top row yellow area).
Figure 2: Filter characteristics of older double dichroic beam splitter compared to AOBS. A spectral band around the nominal filter wavelength is suppressed (white vs. light blue). Fluorescence light passing through the filter is detected (dark blue-green), while some light is lost (yellow). Due to its steep cutoff, the AOBS hardly suffers from this inherent deficiency of dichroic beam splitters.
A solution to this dilemma is the use of a filter-free concept for beam splitting. With the acousto-optical beam splitter (AOBS) Leica has introduced a beam splitting technology which simultaneously acts as a multiplexer for several wavelengths (Figure 3).
Figure 3: Acousto-optical beam splitter (AOBS). A crystal made of the quartz homolog tellurium oxide (TeO2) acts as a wavelength-specific deflector. To this end, an acoustic wave in the kHz range is coupled into the crystal thus changing its dispersion properties. This principal can be applied to up to eight wavelengths at once through superposition of several acoustic frequencies. The underlying physical principle is based on deformations in the crystal lattice including its electrons. They interact with the electric field of incident light and therefore guide it differently depending on the electron’s respective displacement.
In practical terms, particularly in multi-spectral imaging, the steep cutoff of the AOBS provides a more efficient beam splitting by permitting more light to be detected (Figure 4). Moreover, using dichroic mirrors, certain dye combinations such as a triple label of GFP, YFP and mCherry cannot be recorded simultaneously. No such filter combination exists, hence, limiting the experimental flexibility to freely combine any fluorescent marker (Figure 5A). The AOBS’s steep cutoff again helps to address this requirement by allowing even very narrow emission bands to be recorded in parallel, backed by the sensitive detection system of the Leica TCS SP8 (Figure 5B). A live cell example of GFP, YFP and mCherry being simultaneously recorded is provided here. Even more combinations are possible because the AOBS can be reprogrammed to a different wavelength within microseconds and even simultaneously split up to 8 laser lines. So, line-sequential scanning provides more pseudo-simultaneous multi-channel recording on top of the simultaneous ones described above. Sequential scanning is not possible with all dichroic filter combinations in contrast. These properties make the AOBS ideal for multi-laser setups where 8! = 40320 different combinations are possible. This applies all the more for the white light laser (WLL2) where literally trillions of excitation combinations result and the idea of a fully tunable white confocal has been realized.
Figure 4: Light efficiency of the AOBS. Top row: excitation (dotted lines) and emission spectra (solid lines) in multi-color samples with Cy2 (green), Cy3 (yellow) and Cy5 (red), respectively. Area fills represent transmitted fluorescence at > 90% transmission of dichroic beam splitters (A) and the AOBS (B). The detectable fluorescence is integrated and expressed in relation to the integrated emission spectrum of the respective dye (C). The AOBS transmits a larger portion of the fluorescence light (C, red bars) compared to dichroics (C, blue bars).
Figure 5: Flexibility and speed through the AOBS. GFP, YPF and mCherry spectra are shown in green, yellow and red, respectively, as in Figure 4. No combination of dichroic mirrors can record the strongly overlapping emission spectra of GFP and YFP simultaneously. The overlap of typical dichroic wavelength bands is shown in the dashed area fill leaving no room for the necessary 514 nm excitation. In contrast, the AOBS captures both spectra in parallel by allowing for very narrow detection bands in conjunction with the SP detector and optionally with HyDs.