Thorlabs' Ø1/2", Ø1", and Ø2" Pellicle Beamsplitters, which are available with models covering wavelengths from 300 nm - 5 µm, are ideal for use in applications where chromatic dispersion must be minimized (i.e., cases where focused beams are necessary). Pellicle beamsplitters virtually eliminate ghosting since the second surface reflection is superimposed on the first one. However, they are extremely fragile due to the nitrocellulose membrane being only a few microns thick. For each size, Thorlabs offers both coated and uncoated versions. When using a coated optic, light should be incident upon the coated surface first (Note: the coated side of these beamsplitters is the side without the engraving). When using an uncoated optic, light can be incident on either surface.
The curves pictured below show each of the different coating types as well as the uncoated beamsplitting ratios over a variety of wavelengths. Data was obtained for unpolarized, S-polarized, and P-polarized light incident at 45°. In general, P-polarized light will be transmitted more than S-polarized light, and sinusoidal oscillation is present, resulting from thin film interference effects. See the Tutorial tab for more information about interference effects.
Please note that the size of the pellicle beamsplitter (e.g., Ø1") refers to the inner diameter of the aluminum frame, not the outer diameter. Hence, these beamsplitters are not compatible with our standard Ø1/2", Ø1", and Ø2" optic mounts. Refer to the Specs tab for a complete table of dimensions.
Thorlabs Lab Fact: Beamsplitter Package Matters
We present laboratory measurements of the polarization angle, split ratio, and total throughput power of a beam transmitted through Thorlabs plate, cube, and pellicle beamsplitters. While all non-polarizing beamsplitters function similarly, the exact performance is different for different types of beamsplitter. Each type of beamsplitter contains its own advantages and disadvantages compared to other types of beamsplitters. Appropriate choice of beamsplitter is essential to sensitive experimental systems. We present a complete analysis and comparison of optical parameters for three common types of non-polarizing beamsplitters.
For our experiment we used the HRS015 stabilized HeNe as the light source for our investigation. A linear polarizer is used to set the laser beam's polarization axis to 45° in order to provide equal S- and P-polarized light incident on the beamsplitter. The beamsplitter under investigation was then placed in the beampath, and its split beams directed to appropriate detectors. The total power though the optic, polarization states, split ratios, and angle of incidence effects were investigated under this configuration.
The plots below summarize the measured results for all three types of beamsplitters. From these graphs the performance of each optic can be easily compared to one another. The bottom left plot summarizes the results for the total power throughput for each optic. The total power throughput is measured as the fraction of input power. While the plate and pellicle beamsplitters perform rather similarly, the cube shows signs of absorption inside the optic. Additionally, this plot shows the relative insensitivity of throughput power to angle of incidence. The bottom middle graph summarizes the results for the output polarization angle for each optic. The cube shows the most similar polarization angles between the reflected and transmitted beams, with the plate producing the largest difference in polarization between beams. The bottom right plot summarizes the results for the split ratio, as a fraction of input power, for the beamsplitters. Here it can be shown that the plate beamsplitter demonstrates the most ideal for 50/50 power splitting. For details on the experimental setup employed and the results summarized here, please click here.
Beamsplitter Selection Guide
Pellicle Beamsplitters - Pellicle beamsplitters are the best choice when dispersion must be kept to a minimum. They virtually eliminate multiple reflections commonly associated with thicker glass beamsplitters, thus preventing ghosting. In addition, unlike plate beamsplitters, there is a negligible effect on the propagation axis of the transmitted beam with respect to the incident beam.
Pellicle beamsplitters have two disadvantages: They exhibit sinusodial oscillations in the splitting ratio as a function of wavelength, due to thin film interference effects. Click Here for more details. They are also extremely delicate. Since they are fabricated by stretching a nitrocellulose membrane over a flat metal frame, the beamsplitter cannot be touched without destroying the optic. Thorlabs offers pellicle beamsplitters mounted in metal rings for use in kinematic mounts as well as 30 mm cage cube-mounted pellicles.
Polarizing Beamsplitters - Thorlabs’ polarizing plate and cube beamsplitters split randomly polarized beams into two orthogonal, linearly polarized components (S and P), as shown in the diagram to the right. S-polarized light is reflected at a 90° angle with respect to the incident beam while p-polarized light is transmitted. Polarizing beamsplitters are useful in applications where the two polarization components are to be analyzed or used simultaneously. Thorlabs offers broadband 16 mm cage cube-mounted, broadband 30 mm cage cube-mounted, and broadband unmounted polarizing beamsplitter cubes, as well as laser line 30 mm cage cube-mounted and laser line unmounted cubes. Additionally, Thorlabs offers wire grid polarizing beamsplitters which have a larger Angle of Incidence and work with uncollimated light. For applications requiring higher power, we also offer high-power polarizing beamsplitting cubes.
Non-Polarizing Beamsplitting Cubes - These cubes provide a 50:50 splitting ratio that is nearly independent of the polarization of the incident light. The low polarization dependence of the metallic-dielectric coating allows the transmission and reflection for s- and p-polarization states to be within 10% of each other. These beamsplitters are particularly useful with randomly polarized lasers and are specifically designed for applications in which polarization effects must be minimized. Thorlabs offers 16 mm cage cube-mounted, 30 mm cage cube-mounted, and unmounted beamsplitter cubes.
Plate Beamsplitters - Thorlabs' plate beamsplitters are optimized for an incidence angle of 45° and feature a dielectric coating on the front surface for long-term stability. To help reduce unwanted interference effects (e.g., ghost images) caused by the interaction of light reflected from the front and back surfaces of the optic, a wedge has been added to the round versions of these beamsplitters. Dispersion, ghosting, and shifting of the beam may all be potential problems, however. These are the best choice for a general-purpose beamsplitter. Thorlabs offers both polarizing and nonpolarizing plate beamsplitters.
Economy Beamsplitters - These are the most cost effective of all the beamsplitter types. Thorlabs' economy beamsplitters, which have an exposed oxide coating on one side and are uncoated on the other side, are designed to have either a 50:50 or 30:70 splitting ratio throughout the visible spectrum (450 - 650 nm) when used with unpolarized light incident at 45°.
Please note that the Fresnel reflections off of the uncoated back surface of these economy beamsplitters can lead to interference effects in the reflected beam. For applications sensitive to these effects, consider using a beamsplitting cube or a pellicle beamsplitter.
Polka Dot Beamsplitters - This type of beamsplitter consists of a glass substrate with a vacuum-deposited reflective coating that is applied over an array of apertures, giving the beamsplitter a "polka dot" appearance. Half of the incident beam is reflected from the coating, and half of the beam is transmitted through the uncoated portion of the substrate.
Polka dot beamsplitters are useful over a wide wavelength range and are negligibly angle sensitive, which makes them ideal for splitting the energy emitted from a radiant source. These are not recommended for imaging applications, such as interferometry, as the polka dot pattern will affect the image.
Thin-Film Interference Effect
The plots below, which show transmission as a function of wavelength, depict a sinusoidal fluctuation that is caused by interference effects. These effects will occur for all pellicle beamsplitters and are sometimes averaged out when data is displayed.
The graph to the right depicts the measured reflectance of an 8:92 beam sampler without averaging the sinusoidal oscillations that result from thin film interference. The frequency and amplitude of the pattern depends several factors, such as the thickness of the film, the thickness of any coating present, the angle of incidence of the incoming light, the polarization of the incoming light, and the bandwidth of the light incident on the pellicle.
What is a Thin Film?
For the pellicle beamsplitters featured here, the information under the Specs tab states that nfilm = 1.5 at 550 nm. Therefore, for this incident wavelength, the wavelength in the pellicle membrane itself is
Thorlabs offers both fixed and kinematic mounting options for our Ø1/2", Ø1", and Ø2" pellicle beamsplitters. Two 2-56 capscrews included with the mount are used to attach the pellicle. Pellicle mounts are post-mountable via 8-32 or M4 x 0.7 tapped mounting holes (KM05BP and KM05BP/M), or #8 (M4) counterbores (all other pellicle mounts).