We maintain a stock of premium quality interference filters.  Please check our page of Interference Filters for current availability and spectra.  The filters used in our instruments are custom designed to our high standards and are not available elsewhere as "stock" items.  These filters have very high wavelength specificity.  Typical transmission in the out-of-band region is > 6 OD. 

Incident light is passed through two coated reflecting surfaces. The distance between the reflective coatings determines which wavelengths destructively interfere and which wavelengths are in phase and will ultimately pass through the coatings. If the reflected beams are in phase, the light is passed through two reflective surfaces. If, on the other hand, the multiple reflections are not in phase, destructive interference reduces the transmission of these wavelengths though the device to near zero. This principle strongly attenuates the transmitted intensity of light at wavelengths that are higher or lower than the wavelength of interest.

In an interference filter, the gap between the reflecting surfaces is a thin film of dielectric material called a spacer. It has a thickness of one-half wave at the desired peak transmission wavelength. On either side of this gap are the two reflecting layers. The reflecting layers actually consist of several film layers, each of which is a quarterwave thick. This sandwich of quarterwave layers is made up of an alternating pattern of high and low index material, usually zinc sulfide and cryolite, respectively. Together, the quarterwave coatings forming the reflective layer is called a stack. The combination of two stacks and the spacer comprise a one cavity bandpass filter. This is shown schematically in Figure 1, below. The number of layers in the stack is adjusted to tailor the width of the bandpass.

Figure 1: Single cavity bandpass filter

In practice, a single cavity bandpass filter does not exhibit a sharp transition between the passband and out-of-passband wavelengths. To sharpen this cutoff, it is common practice that several cavities are layered sequentially into a multicavity filter design. A multicavity design also dramatically reduces the transmission of out-of-band wavelengths.

Figure 2, below, illustrates the many coating layers that make up a three cavity bandpass filter. Each separate cavity is separated by a coupling layer. Numerous filters having a three cavity design are commercially available. However, for discriminating in a three cavity design, [missing text].

Figure 2: Three cavity bandpass filter

To complete the interference filter, another set of thin-film coatings are applied to the second substrate to block the transmission of wavelengths that are further away from the passband of interest. This blocking layer is essential in filters used for the measurement of fluorescence. It prevents "shoot through" of undesired wavelengths from the illumination source to the detector. The blocking layers and passband layers are held together in a protective metal case using optical epoxy. This is illustrated below, in Figure 3.

A third factor that must also be considered is time, or more properly, temporal resolution. If your application requires you to monitor a rapidly changing event, it may not be possible to collect sufficient photon counts in the time available to record the event with high fidelity. That is, the time for each measurement period is reduced as the frequency of measurement periods is increased.

These three factors can be considered as opposing forces in your quest to obtain meaningful data from your fluorometer. The interrelation between these factors is readily apparent if they are viewed as vertices in a triangle, as shown in Figure 4, below.

Figure 4: Balancing between sensitivity, wavelength selectivity and temporal resolution in fluorescence measurements

Arranged in this manner, it is easy to see the relationship between these three parameters. One of the three factors must be sacrificed when the other two are optimized. For instance, if temporal considerations are not an issue, data can be acquired at a slow rate with high wavelength selectivity (i.e., narrower passbands). Collecting light at a lower frequency allows for high sensitivity. As another example, consider the case when measuring the fluorescence of a single dye that exhibits a large Stoke's shift. Under these conditions, wavelength selectivity can be sacrificed by using wide passband filters in both the excitation and emission optics to increase the sensitivity, even at a high sampling frequency.

With these factors in mind, one can easily apply some simple rules in choosing which interference filters are best for a given application. The first two items to consider are 1) the required time resolution and 2) the Stoke's shift of the fluorescence species. These two parameters are generally set by the event which you wish to monitor and the given fluorescent dye used in making the measurement, as described below.

Time resolution. Does your application require the measurement of fluorescence at a high sampling rate? The general rule of thumb is that, in order to reproduce the observed signal with reasonable fidelity, you should sample at a 10-times the rate of change in the signal. For instance, to accurately capture and then represent calcium oscillation in a cell with a calcium-sensitive fluorescent indicator occurring at a frequency of 5 Hertz, one should measure this event at a minimum frequency of 50 Hertz, or one sample every 20 milliseconds. In this instance, a high sampling frequency sets a constraint on the other two variables of sensitivity and wavelength selectivity. Thus, if this sampling frequency is required, the bandpass of the excitation and/or emission wavelengths may have to be increased to acquire sufficient signal.

Stoke's Shift. The Stoke's shift is the difference in nanometers between the peak excitation and emission wavelengths of a fluorescent species. Some fluorescent dyes, such as FITC, exhibit a small Stoke's shift, whereas others, such as fura-2, exhibit a large Stoke's shift. The following graph, Figure 5, exhibits hypothetical fluorescence spectra of a dye which exhibits a small Stoke's shift. The Stoke's shift in this example is only 30 nm.

The small Stoke's shift of this dye will limit the usable bandpass. Ideally, the center wavelengths for the setting of the excitation and emission passbands should be at the peak values of the spectra. As can be seen, this will be difficult in this example without causing overlap between the excitation and emission bandpass. In this instance, either the entire excitation spectra cannot be used to excite the dye or all the light from the emission of the dye cannot be collected for detection. Thus, either time resolution or sensitivity must be sacrificed in this instance in order to achieve high wavelength selectivity.

Figure 6, below, illustrates the transmission of several hypothetical interference filters having a 10 nm bandpass at a center wavelength of 450 nm. This figure shows the typical transmission profile as a function of wavelength for filters constructed with different numbers of cavities. Note that as the number of cavities used in the construction of the filter is increased, the degree of attenuation of light in the out-of-band is greater at wavelengths further away from the center wavelength. With a high number of cavities, the transmission of the interference filter begins to approach an ideal "square wave" appearance in which there is a very sharp transition between the transmission of in-band and out-of-band wavelengths. In the design of a fluorometer, this translates to the ability to achieve higher wavelength discrimination and higher energy throughput, while also retaining the ability to measure the fluorescence of dyes that exhibit small Stoke's shifts.

To further illustrate this concept, assume that one wished to measure the fluorescence of a species using 450 and 480 nm as the excitation and emission wavelengths using interference filters with center wavelengths at these values. The following figure, Figure 7, shows the hypothetical transmission values for these filters using 3 or 5 cavity designs. Note that with use of the 3 cavity filters, the excitation filter transmits approximately 1% (i.e., OD 2) of the peak excitation energy at 470 nm. The emission filter also transmits this wavelength to the same extent. This results in what is termed "bleed through". This is excitation energy which can be passed directly through to the detector. The actual extent of bleed through can be minimized by the design of the sample chamber or fluorescence microscope, but this should be avoided by proper selection of filters. The signal resulting from this effect is added to the actual fluorescence of the sample. The added signal increases the background fluorescence and severely limits the lower limit of fluorescence detection.