TIFR Microscopy

Compare TIRF Geometries

Total Internal Reflection Fluorescence (TIRF) has established itself as an indispensable tool for single molecule detection, super resolution microscopies, lipid rafts studies, analysis of biomolecular interactions, and other areas of life sciences [1-8, 11-15]. In particular, TIRF is “…a method uniquely suited to image the plasma membrane with its associated organelles and macromolecules in living cells. The method shows even the smallest vesicles made by cells, and can image the dynamics of single protein molecules” [16].

TIRF selectively excites a ~100-nm layer of the specimen next to the surface, while the bulk of specimen is not illuminated and, respectively, does not fluoresce. This spatial selectivity (in comparison, the confocal scheme excites ~1,000 nm) allows for providing the contrast sufficient for the detection of single molecules. Since single molecules are smaller than the wavelength, this allows to investigate the architecture and functions with resolution better than the diffraction limit.

There are three popular geometries that implement the TIRF effect: (i) prism-, (ii) lightguide-, and (iii) objective-based geometries, as shown in the schemes above. Each geometry has its own set of advantages and limitations. This web page analyzes errors caused by the stray light in TIRF microscopy and its interference with the TIRF effect. It compares the advantages and limitations of prism-, lightguide- and objective-type TIRF microscopy, discusses the sources of stray light in each of the geometries, and suggests methods of stray light mitigation. This web page will help you choose the TIRF geometry that is optimally suited for your studies.

TIRF Principles. The phenomenon of total internal reflection occurs at the interface between two optical media with different refractive indices, e.g. glass/water. If the angle of incidence is greater than critical, the incident light reflects back into the glass and generates a profile of exponentially decaying intensity at the interface, termed the Evanescent Wave (EW) (Fig. 1). The critical angle is given by eq. (1) (Fig.1). For the glass/water interface the angle is 62 degrees. EW intensity is maximal at the surface and decays exponentially with distance – eq. (3) (Fig.1) The depth of penetration depends on the wavelength, the angle of incidence, and the ratio of refractive indices eq. (2).  For a typical TIRF experiment with a glass/water interface, the depth of penetration is ~100 nm.

TIRF Theory and Practice. In theory, the intensity of the EW exponentially decays with distance, as shown in Fig. 1. In practice, however, autofluorescence in lenses and/or other optical elements, scatter, reflections, and refractions produce undesirable rays of light, collectively termed “stray light.” Stray light contaminates the exponential decay of the EW, excites the bulk of specimen, and deteriorates the TIRF effect, as shown in Fig. 2. This is an especially serious problem in the case of objective-type TIRF.

Sources of Stray Light. All optical materials, to a certain extent, auto-fluoresce and/or scatter light, and all surfaces and interfaces between optical parts reflect, refract, and scatter light. Due to the combination of these factors, the undesirable stray light is present in all practical systems. The intensity of stray light and its interference with the TIRF effect depend critically on the optical scheme. In certain cases, its intensity is too large to be neglected.

TIRF Geometries. There are three popular geometries in TIRF microscopy: through-objective,  prism-, and lightguide-based optical schemes (o-TIRF, p-TIRF, and lg-TIRF, respectively). For a number of reasons, the companies of Nikon, Olympus, Zeiss, and Leica have aggressively marketed only the o-TIRF geometry. Objective-TIRF employs expensive specialized high NA objectives and other sophisticated optics; for this reason, a typical o-TIRF microscopy system costs ~$80,000 or more. Ironically, it has been reported in the literature that o-TIRF has the worst signal-to-background ratio. To make matters worse, it is too rigid; it can be used only with specialized TIRF objectives.  Besides this, there are more cost-efficient TIRF geometries (~$10k) that demonstrate superior signal-to-background ratios, namely p-TIRF and lg-TIRF. However, they were not offered commercially, until recently, in 2010, when TIRF Labs started to market them. In the past, each research group built p-TIRF and lg-TIRF systems on their own. Since 2010, our customers generated unique TIRF data and demonstrated superior advantages of p- and lg-TIRF geometries on a number of applications, including single molecule detection [1, 2]. The sections below compare the geometries, analyze sources of stray light, and outline ways to mitigate the stray light interferences.

Objective-type TIRF Geometry. Fig. 3 shows the schematics of the o-TIRF system.  The main  feature, which might appear elegant at first glance, is the use of the emission path for delivering the excitation light to the glass/water interface. This feature has been acquired from epi-fluorescence. The o-TIRF scheme uses large angles of incidence, greater than the critical angle for glass/water interface, which is 63 degrees. Microscope objectives with a Numerical Aperture (NA) smaller than 1.38 do not support such angles. Therefore, the o-TIRF scheme relies on specialized high-NA objectives.

Sources of stray light in o-TIRF. Fig. 3 illustrates numerous potential sources of stray light in o-TIRF. Not surprisingly, significant interferences of stray light have been reported for the o-TIRF geometry [3-8].The intensity of stray light at the surface amounts to 10-15% of that of the evanescent wave; the ratio increases exponentially with the distance. Fluorophores in the bulk of the specimen are excited and the TIRF effect is compromised. In many instances, the intensity of stray light changes unpredictably. Two major sources of stray light were identified as originating from: (i) the TIRF objective, and (ii) the rest of the microscope optics [4, 5].Only minor contributions were detected due to the scatter at the TIRF glass/water interface and at refractive-index boundaries within the specimen, including live cells [4, 5].

The first group of sources (i) is related to undesirable autofluorescence, scatter, and reflections inside the objective. The intense excitation light travels through multiple lenses and interfaces on its way to and from the TIRF surface (Fig. 3). The quality of the optical glass – which should be minimally fluorescent and scattering – and the surface quality are critically important to the minimization of the intensity of stray light. To our knowledge, the systematic comparison of TIRF objectives from the standpoint of the intensity of stray light has not been performed yet. Our own tests on a small number of TIRF objectives, our analysis of the literature, and reports from our customers and colleagues indicate that all TIRF objectives demonstrate a significant intensity of stray light due to autofluorescence and scatter. The autofluorescence and scatter of the front lens, which is in direct contact with the specimen, appears to be the most contributing factor.  

The second group of stray light (ii) originates from the optics inside the microscope. A significant amount of stray light is generated at the dichroic mirror. Even a high-quality dichroic mirror scatters and transmits a certain portion of light, which, in a perfect world, would be ideally reflected and blocked. The leaking of excitation light through the dichroic mirror, as well as through the emission filter, results in an increased background signal when it arrives at the photodetector.

In o-TIRF, the intensity of stray light changes unpredictably with the angle of incidence and XYZ coordinates. It increases with the amount of imperfections located on the path of the excitation light. Certain types of imperfections are distributed randomly, while other types exhibit more systematic patterns of their occurrence.  If the angle of incidence increases, the depth of penetration and the intensity of the EW excitation decreases, while the average intensity of stray light remains the same or increases. If you are performing variable angle TIRF experiments using o-TIRF geometry, the effect of stray light should be carefully taken into account. 10-15% of stray light at the surface is typical for a high quality optical system [7, 4, 5]. In certain cases, the intensity of stray light is even larger – comparable to that of the EW. In such cases, the depth of penetration calculated using eq. (3) (Fig.1) does not describe the intensity profile anymore and can mislead the interpretation of biological TIRF images [3-8]. In fact, since the EW decays exponentially, the error caused by stray light increases exponentially with the distance.

 Dichroic mirror and emission filter. The dichroic mirror is the central element of o-TIRF geometry; its quality is critically important for the objective-type TIRF. The excitation light reflected from the dichroic beamsplitter must be focused at the back focal plane of the objective. A significant focal shift or a change in the focal spot size caused by a bend in the dichroic mirror can make it difficult to achieve TIRF, especially if the microscope has a limited ability to adjust the collimation of the excitation beam. Chroma and Semrock made significant progress in improving technical performance of dichroic and bandpass filters. The companies have increased the thickness of dichroic mirrors to keep their flatness, which is necessary for precision focusing in o-TIRF, minimized surface roughness, and the density of pinholes [9, 10]. Even still, the improved filters are not perfect.

 o-TIRF stray light mitigation. First, it appears to be rational to explore the opportunity of using alternative TIRF geometry. If your study dictates the use of the o-TIRF scheme, select a TIRF objective with the smallest amount of stray light. Use a 405 nm laser pointer for a quick assessment and visualization of autofluorescence and scatter in TIRF objectives and coverslips. Always use safety goggles and longpass filters to inspect the objective and coverslips. Select the best quality objectives, coverslips, dichroic mirror and the emission filter. Use an additional excitation filter to block undesirable lines in the excitation light. Make sure that the dichroic mirror, emission filter, and other accessible optical parts are free from dust particles and other contaminations.

Prism-based TIRF. p-TIRF geometry has been shown to provide the best signal-to-background ratio [1-3]. p-TIRF is a geometry where the excitation and emission channels are naturally independent, as it is illustrated in Fig. 4. This separation of lightpaths provides evident advantages to TIRF. Since the excitation light travels through the prism, reflects at the glass/water interface, and escapes through the opposite facet of the prism – the mainstream of the excitation stays away from the emission channel. Indeed, potential sources of stray light are a few reflections at the prism facets, prism/slide interface, autofluorescence and scatter in the prism and slide/coverslip.  However, these few sources are located away from the EW region and minimally affect the EW and minimally enter into the detection channel. Only a small portion of stray light reaches and contaminates EW and enters the emission channel. Crisp, high-contrast TIRF images have been reported for p-TIRF [1-3, 8].

p-TIRF stray light mitigation. If your application permits to enclose your specimen into a closed flow cell, p-TIRF geometry is the best choice for your studies.   In the p-TIRF scheme, the intensity of stray light is negligibly small, provided that high-quality optical materials were selected to minimize the scatter and autofluorescence inside the prism and the TIRF slide. For demanding  experiments, the surface of the TIRF prism and TIRF slide should be of high quality to prevent scatter. At the emission channel, the selection of an efficient emission filter, which blocks the stray light, will reduce the background.

 There are many different configurations of prism-TIRF for inverted and upright microscopes. Contact TIRF labs for details. For example, for live cell studies an upright microscope with a prism-down geometry and water-immersion objective is a geometry well-suited for use with open perfusion chamber. It is difficult to arrange open dishes for working with live cells an inverted microscope. However, several versions of p-TIRF geometry with open perfusion chambers on inverted microscopes is also available from TIRF Labs. These versions of use objectives with working distance  2 mm or more. Contact TIRF Labs for more information info@tirf-labs.com.

Lightguide-based TIRF geometry. Lg-TIRF is a sensible alternative to o-TIRF. It is a flexible geometry available with open perfusion chambers on inverted microscopes, and closed flow cells [11]. Lg-TIRF can be used with dry, water-, and oil-immersion objectives, with UV and visible excitation. The novel method of Shallow Angle Fluorescence Microscopy, in which objects a few microns away from the surface are illuminated,  can be implemented at the same platform [12]. Lg-TIRF principles are close to that of p-TIRF: the excitation light is naturally independent from the emission channel. In lg-TIRF, the excitation light undergoes multiple reflections from the top and the bottom surface of glass or silica coverslip, which serves as the TIRF lightguide. The excitation light escapes from the opposite end of the coverslip. Unlike in o-TIRF, it does not enter the emission channel. The optical scheme of lg-TIRF allows for the use of Petri dish chambers for experiments with live cells. A potential source of scatter and autofluorescence is the coverslip material. Additionally, the top, bottom, and end surfaces of the coverslip play an important role in lg-TIRF. For demanding TIRF experiments, these surfaces should be polished to the highest quality.  

Lg-TIRF stray light mitigation. To minimize the intensity of stray light and to maximize the signal-to-background ratio in lg-TIRF, it is recommended to use coverslips made from high-quality synthetic silica,  the autofluorescence of which is low. The use of coverslips polished to the highest  surface quality reduces scatter from the surfaces. However, regular flame-polished coverslips work well for many lg-TIRF applications. Higher quality coverslips as well as superior emission filters are necessary only for demanding TIRF applications.

In summary, it appears that using the emission channel to deliver the TIRF excitation results in too large of an intensity of stray light in the case of  o-TIRF. Indeed, p-TIRF and lg-TIRF geometries provide superior signal-to-background ratios, because the excitation light is naturally separated from the emission channel. lg-TIRF performs all the functions that o-TIRF does, and can even be used with UV excitation. If your applications involve TIRFing living cells in open chambers, lg-TIRF is the optimal system for this purpose. If you are TIRFing single molecules with bright fluorophores, an inexpensive  objective 1.20 NA will be sufficient. The use of expensive >1.4 NA lenses will increase the amount of light collected by the objective, while the lg-TIRF geometry gives you the freedom to select an objective that is better fitted for your specific task. Visit www.tirf-labs.com/lightguide-TIRFM for additional information.

CCD camera. The TIRF signal is inherently low, because it is generated by a thin layer of the specimen ~0.1 um. Therefore, a low light camera is a necessary module for TIRFing. An electron-multiplying CCD camera is a rational choice. Scientific sCMOS cameras, and low light  CMOS cameras are sufficiently sensitive for many TIRF applications, including single molecule detection [13].

UV-absorbing-blue-light-emitting fluorophores. p-TIRF and lg-TIRF systems can be used with UV excitation, a feature not available in o-TIRF. There is a new class of fluorescent probes that absorb UV light and emit in the range of 360-450 nm.  These fluorophores, including modified nucleic and amino acids, are useful for many lifescience applications. p-TIRF and lg-TIRF systems are well suited for TIRFing UV fluorophores.

 Selecting your microscope objective. Unlike o-TIRF, which can be used only with specialized TIRF objectives, p-TIRF and lg-TIRF systems can be used with high and low NA, dry, water- and oil-immersion lenses, UV or visible light. Many TIRF applications, e.g. real-time microarrays, deal with relatively large amount of fluorescence emitted from relatively large area; dry objectives with low magnification are suitable for such studies. For low light applications, such as single molecule detection and super resolution methods, objectives with high NA provide better sensitivity, because they collect more light (equation (4)).  When you are choosing the objective, take into account that the efficiency of collecting light, G, is proportional to the square of the numerical aperture of the objective, NA,  and inversely proportional to the square of the magnification, X:

                                                                      G = k S  (NA)2 (X)-2    (4)

where k- is a coefficient, constant for the given microscope. The table below enumerates the relative efficiency of objectives with different NA and X values. For example, a 60X/1.40 objective, which is available as a cost-efficient oil-immersion lens, provides 87% efficiency, relatively inexpensive to the costly 60X/1.49 lens. The light-collection efficiency of the 60X/1.40 lens 2.4 times exceeds that of 100X/1.49 objective.

Table. Relative light-collecting efficiency of microscope objectives with different values of the numerical aperture (NA) and magnification (X).

Magnification / Numerical Aperture	60X/ 1.49	60X/ 1.4	63X/ 1.2	60X/ 1.0	100X/ 1.49	100X/ 1.2	40X/ 0.80
Relative efficency of collecting light	100%	87%	58%	45%	36%	23%	65%

The proximity of a fluorophore to a dielectric surface results in an anisotropy of emitted fluorescence [14, 15]. If a fluorophore is in aqueous phase in close proximity to an interface with glass, it does not emit light isotropically, unlike fluorophores in the bulk of the solution. In the case of TIRF, emission is produced in a complex anisotropic pattern that depends on the orientation of the fluorophore transition dipoles with respect to the surface, dielectric properties of the media, and specific binding to biological structures. An example of TIRF emission anisotropy is shown in Figure 6 below for molecules oriented perpendicular and parallel to the interface between silica and water (Ref: “olympus.magnet.fsu.edu/primer/techniques/fluorescence/tirf/tirfintro.html”).

In summary, the problems of stray light, a poor signal-to-background ratio, and a compromised quality of the evanescent wave are inherent only to the objective-TIRF geometry. Significant interferences of stray light with TIRF effect and undesirable excitation of the bulk of the specimen have been reported only for the case of the objective-TIRF geometry [1-6]. These interferences are caused by the fact that the excitation and emission channels share the same optics. On the other hand, prism-TIRF and lightguide-TIRF geometries demonstrate an excellent quality of the TIRF effect, a minimum amount of stray light, and crisp and high-contrast TIRF images. Both p-TIRF and lg-TIRF geometries employ optical schemes in which the excitation light is independent from the emission channel. Published articles show that p-TIRF and lg-TIRF geometries are well-suited for single molecule detection, living cell membrane studies, and other areas of lifesciences. For supersensitive TIRF experiments, it is rational to use p-TIRF and lg-TIRF with microscope objectives with high or moderate NA numbers (>1.0 and more) to ensure that the objective collects a sufficient portion of the emitted fluorescence. For additional information please contact TIRF Labs: info@tirf-labs.com.

Lightguide-based TIRF geometry. Lg-TIRF is a sensible alternative to o-TIRF. It is a flexible geometry available with open perfusion chambers on inverted microscopes, and closed flow cells [11]. Lg-TIRF can be used with dry, water-, and oil-immersion objectives, with UV and visible excitation. The novel method of Shallow Angle Fluorescence Microscopy, in which objects a few microns away from the surface are illuminated,  can be implemented at the same platform [12]. Lg-TIRF principles are close to that of p-TIRF: the excitation light is naturally independent from the emission channel. In lg-TIRF, the excitation light undergoes multiple reflections from the top and the bottom surface of glass or silica coverslip, which serves as the TIRF lightguide. The excitation light escapes from the opposite end of the coverslip. Unlike in o-TIRF, it does not enter the emission channel. The optical scheme of lg-TIRF allows for the use of Petri dish chambers for experiments with live cells. A potential source of scatter and autofluorescence is the coverslip material. Additionally, the top, bottom, and end surfaces of the coverslip play an important role in lg-TIRF. For demanding TIRF experiments, these surfaces should be polished to the highest quality.  

Lg-TIRF stray light mitigation. To minimize the intensity of stray light and to maximize the signal-to-background ratio in lg-TIRF, it is recommended to use coverslips made from high-quality synthetic silica,  the autofluorescence of which is low. The use of coverslips polished to the highest  surface quality reduces scatter from the surfaces. However, regular flame-polished coverslips work well for many lg-TIRF applications. Higher quality coverslips as well as superior emission filters are necessary only for demanding TIRF applications.

In summary, it appears that using the emission channel to deliver the TIRF excitation results in too large of an intensity of stray light in the case of  o-TIRF. Indeed, p-TIRF and lg-TIRF geometries provide superior signal-to-background ratios, because the excitation light is naturally separated from the emission channel. lg-TIRF performs all the functions that o-TIRF does, and can even be used with UV excitation. If your applications involve TIRFing living cells in open chambers, lg-TIRF is the optimal system for this purpose. If you are TIRFing single molecules with bright fluorophores, an inexpensive  objective 1.20 NA will be sufficient. The use of expensive >1.4 NA lenses will increase the amount of light collected by the objective, while the lg-TIRF geometry gives you the freedom to select an objective that is better fitted for your specific task. Visit www.tirf-labs.com/lightguide-TIRFM for additional information.

Literature cited:
Lamichhane R, Liu JJ, Pljevaljcic G, White KL, van der Schans E, Katritch V, Stevens RC, Wüthrich K, Millar DP. Single-molecule view of basal activity and activation mechanisms of the G protein-coupled receptor β2AR. PNAS. 2015 Nov 17;112(46):14254-9.
Lamichhane R, Berezhna SY, Gill JP, Van der Schans E, Millar DP. Dynamics of site switching in DNA polymerase. J Am Chem Soc. 2013 Mar 27;135(12):4735-42.
Ambrose W, Goodwin P, Nolan J. Single-molecule detection with TIRF: comparing signal-to background in different geometries. Cytometry 1999, 36(3), 224.
Brunstein M, Teremetz M, Hérault K, Tourain C, Oheim M. Eliminating unwanted far-field excitation in objective-type TIRF. Part I. Biophys J. 2014; 106(5): 1020.
Brunstein M, Hérault K, Oheim M. Eliminating unwanted far-field excitation in objective-type TIRF. Part II. Biophys J. 2014; 106(5): 1044.
Conibear P, Bagshaw C. A comparison of optical geometries for combined flash photolysis and TIRF microscopy. J Microsc, 2000, 200(3): 218-29.
Mattheyses A, Axelrod D. Direct measurement of the evanescent field profile produced by objective-based TIRF. J Biomed Opt, 2006, 11: 014006A.
Schwarz J, König I, Anderson K. Characterizing system performance in total internal reflection fluorescence microscopy. Meth Mol Biol, 2011, 769: 373-86.
https://www.semrock.com/flatness-of-dichroic-beamsplitters-affects-focus-and-image-quality.aspx
https://www.chroma.com/knowledge-resources
Asanov A, Zepeda A, Vaca L. A novel form of Total Internal Reflection Fluorescence Microscopy (LG-TIRFM) reveals different and independent lipid raft domains in living cells. Biochim Biophys Acta. 2010 Feb;1801(2):147-55. See also references at www.tirf-labs.com/applications.html
Asanov A, Zepeda A, Vaca L. A platform for combined DNA and protein microarrays based on total internal reflection fluorescence. Sensors (Basel). 2012;12(2):1800-15.
Protasenko V, Hull KL, Kuno M. Demonstration of a Low-Cost, Single-Molecule Capable, Multimode Optical Microscope. Chem. Educator 2005, 10, 269282.
Ruckstuhl T, Enderlein J, Jung S, and Seeger S, Forbidden Light Detection from Single Molecules. Anal. Chem. 72(9), 21172123 (2000).
Enderlein J, Ruckstuhl T, The efficiency of surface-plasmon coupled emission for sensitive fluorescence detection. Opt. Express 13(22), 88558865 (2005).
Steyer JA, Almers W. A real-time view of life within 100 nm of the plasma membrane. Nat Rev Mol Cell Biol. 2001, 2(4), 268.