Reflected light allows us to perceive the world around us. Light bounces off objects, enters our eyes, and our brains process the information to allow us to recognize properties such as color, distance and shape. Without these reflections, we would only be able to see objects that emit their own light, such as the sun. In the scientific world, a spectrometer replaces the eye and is used to gather quantitative information.
When light hits a surface, it can be reflected, or transmitted or absorbed. Reflection is the scattering of light back in the direction of the incident light. Transmission occurs when the light passes through a medium and is the reason we can see through windows. Absorption occurs when some of the energy of the light is retained by the sample.
The law of reflection
A ray of light bouncing off the surface behaves in a very predictable manner. The angle of incidence is equal to the angle of reflection: θi = θr, where θi is the angle of incidence and θr is the angle of reflection, as shown in Figure 1. The reflection of light is either specular or diffuse.
Figure 1 – Law of reflection.Specular reflection
This type of reflection takes place when parallel incoming rays of light hit a flat, smooth surface, off which they will reflect in the same direction (Figure 2a). Examples include a reflection in a mirror, an image on an unperturbed lake surface, and the determination of thin-film layer thickness.
Figure 2 – a) Specular and b) diffuse reflection.Diffuse reflection
Diffuse reflection occurs when a surface is not smooth. If surface roughness is similar to or larger than the light’s wavelength, the rays will scatter in a multitude of directions (Figure 2b). In diffuse reflection, the law of reflection is still valid, but the angle of the surface is changing such that incoming parallel beams hit the different angles of the surface. Diffuse reflection is more commonly used to measure the various characteristics of samples such as powders. In the majority of spectroscopic cases, diffuse reflectance is the measurement of interest.
Parameters to be considered when using diffuse reflectance spectroscopy include sample distance, spot size and surface.
Sample distance
The distance a light beam travels is often determined by the requirements of the sample. For certain measurements, a contact measuring head will give the most consistent data. Other measurements, such as those taken with samples on a moving belt, should be performed from a distance.
Distance determines how much reflected light can be gathered and sent back to the detector: if the light is divergent and the distance from the sample changes, a different amount of light is sent to the detector. This can cause both a baseline shift and spectral artifacts. Baseline shifts can be caused by a change in the amount of overall amount of light that hits the detector. If the distance changes, then more or less light could reach the detector. Since the calculation of the percent reflection (%R) is based on a ratio between a standard and the sample measurement, both at one specific distance, the change in the amount of overall light hitting the detector would cause the full spectrum to shift and the baseline to move. This problem can be alleviated by using a well-collimated beam, which prevents the light from diverging, regardless of the distance. For specular reflection, collimation will make the amount of light hitting the sample consistent with the amount reflecting back. However, for diffuse reflection, with light scattering in different directions, the distance of the measurement head could determine how much light is collected and/or lost.
Figure 3 – Divergent light.Spectral artifacts can occur if the light is divergent. Light transmitted by a fiber-optic cable or other light sources will often diverge unless special care is taken. While the average divergence of the light beam can be calculated, shorter wavelengths will diverge more quickly than longer ones. Thus, if a beam is divergent and there is variation in the sample, a detector may see different levels of varying wavelengths (Figure 3), which creates a spectral artifact.
Spot size
Both the sample and the information needed must be considered to determine the appropriate spot size. Is the sample a large sheet roller or a fine powder? Most often, the entire area of interest is illuminated, though this may not be possible for very large or very small samples.
With large samples, it may not be practical to illuminate the entire sample simultaneously. For example, a conveyor belt containing samples may be several feet wide, making it impossible to collect all of the reflected light. Smaller, representative spot sizes may be chosen, or the measurement spots moved so that most of the sample on the belt can be scanned sufficiently. In these cases, the operator must decide if and how many scanned smaller sample sections adequately represent the whole. Another option is to move the measurement spot(s) to allow more areas of the sample to be measured, and a map of the measurement spots of the sample can be created. Either way, the measurement spot size must illuminate a representative portion of the sample.
For medium-sized samples, spot size is often chosen to be nearly the size of the sample to get the maximum amount of information. The spot should not be larger than the sample, since any light that does not illuminate the sample is generally wasted or contributes to noise in the signal.
With small samples—grains or marble-sized substances—it is important to consider inhomogeneity and packing density. Spot size is usually chosen so that it illuminates several samples, which helps account for the pack density. If the spot size is too small, every data point could be different, since the illuminated portion of the sample could quickly vary between a grain and an empty spot (Figure 4a). Different places on the sample could be measured with radically different values obtained. However, a spot size that is large enough will yield an average of samples and empty spots; thus the signal will not change drastically (Figure 4b). The inhomogeneity of a full sample will also affect the choice of spot size. If the sample is not homogeneous and the sampling size is small (Figure 4c), the measurement may miss the defect and lead to incomplete analysis. A larger spot size would give a better idea of the sample composition (Figure 4d).
Figure 4 – a) Small spot size on a flat, homogeneous surface. b) Large spot size on a flat, homogeneous surface. c) Small spot size on an inhomogeneous surface. d) Large spot size on an inhomogeneous surface.For the smallest samples, such as powders, the desirable spot size can vary. Most powders are well packed and the empty space between granules is insignificant. Furthermore, it can be fairly difficult to create a spot size that is appropriate for measuring a single grain among many. A larger spot size may be indicated, and can provide data for a “large” portion of the powder. The reflected signal level may correlate with the amount of inhomogeneity, allowing a calculation for the amount of other substances.
Surface
The ideal sample surface is a macroscopically flat surface. The material of interest should be spread uniformly within the inspection area to create a homogeneous hemisphere of backscatter (black hemisphere in Figure 5a). As the surface begins to angle relative to the light source, the sphere of backscatter changes (Figure 5b). Therefore, it is much simpler to create a setup so that the light reflecting back to the detector has a constant distance and constant angle. A curved or irregular surface has two main issues: the changing distance and the light detected. The changing distance causes the issues mentioned above. However, the larger issue is that it alters what reflectance is picked up where, and the reflectivity of the sample will seem to change, depending on the angle of illumination and detection. A curved or irregular surface often requires a much more complicated setup to account for the changing macroscopic angles.
Figure 5 – Effects of illuminating a nonflat surface.Distance or contact measurements
Occasionally, due to the factors mentioned above, a question of distance versus contact measurement arises. Distance measurements are useful when samples pass by the measurement head quickly (e.g., a sample traveling on a conveyor belt), or when contact with the sample could damage either the sample or the probe. Most measurement heads are built to work optimally at a set distance, so, while the probe can be set at a fixed distance from the sample, problems arise if the sample is highly porous or has varying transparencies. This can produce the effect of different sample distances and artifacts from the amount of light backscattered. A contact measurement allows measurement at the window of the measurement head, thus eliminating potential depth penetration, although concerns of sufficient sampling exchange need to be considered.
Diffuse reflection setup
Figure 6 – Typical illumination setups of measurement heads from the a) top and b) side views.A common diffuse reflection setup has a reflection probe with six or seven illumination fibers surrounding a single collection fiber. These illumination fibers are oriented so that a large portion of the light overlaps—this is the area that emits the strongest signal. The collection fiber in the center picks up the reflected light from all the positions, but the signal is strongest where the beams overlap (see Figure 6). To avoid specular reflection, the window of the probe can either be angled relative to all the fibers, or the window can be normal to the collection fiber and the illumination fibers are angled with respect to the window.
Conclusion
It is important to consider factors such as sample distance, spot size and surface while performing reflection measurements to ensure reliable analysis and maximize information-gathering.
Additional reading
- Hecht, E. Optics; Black, A., Ed.; Addison-Wesley: San Francisco, Calif.; 4th ed., 2002; ISBN-13: 978-0805385663.
- Lynch, C. (2016, May 23). Business Development Manager / tec5USA.
Stacey Carrier, Ph.D., is a technical sales manager, Hellma USA, Inc., 80 Skyline Dr., Plainview, N.Y. 11803, U.S.A.; tel.: 516-939-0888, ext. 108; e-mail: [email protected]; www.hellma.com