Rayleigh-Sommerfeld Integrals

How does light bend around an obstacle, or spread out after passing through a small opening? This fundamental question, a phenomenon known as diffraction, has challenged physicists for centuries. While early intuitive ideas like Huygens's principle offered a conceptual glimpse, their mathematical formulation by figures like Kirchhoff introduced profound inconsistencies, creating a paradox that undermined the theory's own foundations. This article navigates the journey to a more rigorous understanding of wave diffraction. It addresses the critical flaw in Kirchhoff's theory and introduces the elegant and mathematically sound solution provided by the Rayleigh-Sommerfeld integrals. First, the "Principles and Mechanisms" chapter will delve into the mathematical heart of the problem, contrasting the inconsistent Kirchhoff approach with the self-consistent Rayleigh-Sommerfeld strategy. Subsequently, the "Applications and Interdisciplinary Connections" chapter will showcase the immense power of this theory, revealing its applications not just in optics, but across diverse fields like acoustics and even cosmology. Let's begin by examining the core principles that govern how we mathematically predict the behavior of diffracted waves.

Imagine you are standing on the edge of a calm pond. You toss in a single pebble. A perfect circular wave expands outwards. Now, what if you placed a large board with a small slit in it into the water? The wave hits the board, and a new wave seems to emerge from the slit, spreading out in a completely new pattern on the other side. How can we predict this new pattern?

This is the fundamental question of diffraction, and our first and most beautiful intuition comes from the great Dutch scientist Christiaan Huygens. He proposed that every point on a wavefront can be thought of as a tiny new source, a pebble dropped, creating its own little wavelet. The new wavefront a moment later is simply the grand sum, the combined ripple, of all these tiny wavelets. It's a wonderfully poetic and powerful idea. But as with many beautiful ideas in physics, the devil is in the details. How, exactly, do you sum them?

The first to give this a solid mathematical footing was Gustav Kirchhoff. He took Huygens's idea and, using a bit of powerful mathematics involving Green's identities, created what we now call the Helmholtz-Kirchhoff integral theorem. It was a formula that promised to calculate the wave field at any point behind an aperture, provided you knew the field in the aperture.

To make the formula work, Kirchhoff had to make some "reasonable" assumptions about the light field at the screen. He proposed what we now call the ​​Kirchhoff boundary conditions​​:

1. In the open aperture, the light wave is exactly the same as the incident wave that would be there if the screen were absent.
2. On the dark, opaque part of the screen, and in its immediate shadow, the wave amplitude is simply zero.

This seems perfectly sensible, doesn't it? Where there's a hole, light gets through; where there's a block, it doesn't. But here, we must adopt the physicist's skepticism and ask, "Is that really allowed?"

A wave isn't just a static object; it's a dynamic thing that must obey a law of an undulation, the ​​Helmholtz equation​​. This equation links the value of the wave at a point to how it curves and wiggles through space. Kirchhoff's second assumption—that the field is zero on the screen—implies not only that its value, $U$, is zero, but that its rate of change along the surface must also be zero. For the wave to go from non-zero just before the screen to flat-zero on the screen, its derivative, $\frac{\partial U}{\partial n}$, must also be zero.

So, Kirchhoff's "sensible" assumption secretly demands that we specify both the value of the wave and its slope (normal derivative) on the boundary. A mathematician will tell you this is a serious problem. It's like trying to nail a piece of jelly to the wall with two nails very close to each other. You're over-constraining the problem! The Helmholtz equation is what's known as a second-order differential equation; its solution in a region is uniquely determined if you specify either the values on the boundary (a Dirichlet condition) or the normal derivatives on the boundary (a Neumann condition), but not both!

This isn't just a mathematical nitpick; it leads to a genuine paradox. Consider a completely enclosed volume, like a perfectly black box. If a light source is outside, the field inside should be zero. Kirchhoff's theory, with its assumption that $U=0$ and $\frac{\partial U}{\partial n}=0$ on the walls, correctly predicts a zero field inside. Great! But the mathematical logic used to get there is so restrictive it would imply the field is zero everywhere, even outside the box where the light source is shining! The theory is, in a fundamental way, mathematically inconsistent.

We can see this inconsistency in action. Suppose we use a more rigorous theory, but borrow just one of Kirchhoff's assumptions. In a scenario with a circular aperture, if we assume the derivative $\frac{\partial U}{\partial n}$ is as Kirchhoff says and calculate the field $U$ at the center of the aperture, we get the result $U_{calc} = A(e^{ika}-1)$. Kirchhoff's other assumption, however, states that the field should simply be the incident field, $U_{KBC}=A$. These are clearly not the same! The theory contradicts itself. Similarly, if we start by assuming the field $U$ is as Kirchhoff says, we can calculate the derivative $\frac{\partial U}{\partial n}$ on the opaque part of the screen. We find it is not zero at all; for a half-plane screen, it has a value like $-\frac{U_A}{\pi x}$. The assumptions are at war with each other.

So, how do we fix this? The answer lies in making a more humble, more mathematically honest choice. This is the great contribution of Lord Rayleigh and Arnold Sommerfeld. They understood that you can't have your cake and eat it too; you must choose to specify either the field or its derivative on the boundary, but not both. This leads to two distinct, self-consistent theories.

To get a feel for how this works, we can use a beautiful trick from electrostatics: the method of images.

​​First Rayleigh-Sommerfeld Theory (RS1):​​ This theory is built on what mathematicians call a ​​Dirichlet boundary condition​​. We assume we know the field's value, $U$, on the entire $z=0$ plane (equal to the incident field in the aperture, and zero on the screen). To make this work, the theory constructs a special helper function—a Green's function—that is cleverly designed to be zero everywhere on the infinite plane of the screen. How? It imagines a fictitious "anti-source" or image source, lurking behind the screen like your reflection in a mirror, with opposite phase. The combined field of the real source and this anti-source naturally vanishes on the "mirror" plane. The final diffraction formula, the ​​first Rayleigh-Sommerfeld integral​​, is the result of this elegant setup.

$U_1(P) = -\frac{1}{2\pi} \iint_{\text{Aperture}} U(Q) \frac{\partial}{\partial n}\left( \frac{e^{ikr}}{r} \right) dS$

​​Second Rayleigh-Sommerfeld Theory (RS2):​​ This theory uses a ​​Neumann boundary condition​​. It assumes we know the field's normal derivative, $\frac{\partial U}{\partial n}$, on the boundary plane. Its Green's function is built using a co-operating image source, one with the same phase as the real source. This construction ensures that the derivative of the Green's function is zero on the screen plane. The resulting ​​second Rayleigh-Sommerfeld integral​​ then depends only on the derivative of the field in the aperture.

$U_2(P) = \frac{1}{2\pi} \iint_{\text{Aperture}} \frac{\partial U(Q)}{\partial n} \frac{e^{ikr}}{r} dS$

Both of these approaches are mathematically sound. They avoid the over-specification that plagued Kirchhoff's theory. If we return to our black box scenario, both RS1 and RS2, when applied correctly with their respective Green's functions, self-consistently predict a zero field inside, without making nonsensical claims about the field outside.

Naturally, you ask: "We have two self-consistent theories. Which one is 'right'? And do they give the same answers?" This is an excellent question. Physical reality, after all, is just one thing.

The first surprise comes when we look back at the old, flawed Kirchhoff theory. It turns out that Kirchhoff's formula is nothing more than the exact arithmetic mean of the two Rayleigh-Sommerfeld integrals!

$U_K = \frac{1}{2} (U_{RS1} + U_{RS2})$

Kirchhoff's "reasonable" but inconsistent approach was, by a wonderful coincidence, an unwitting compromise between two perfectly consistent theories! This explains why Kirchhoff's theory, despite its flaws, often gives remarkably accurate results: it's the average of two very good theories.

So, do RS1 and RS2 differ? Yes, in their exact mathematical forms, they do. But in the realm where we almost always use them—observing the diffraction pattern many wavelengths away from the aperture—their predictions become virtually indistinguishable. In a typical optical setup (the ​​paraxial approximation​​), we find that the two integrals give essentially identical results. For all practical purposes, in the world of far-field optics, they agree. The choice of which one to use often comes down to mathematical convenience for the problem at hand.

This entire discussion might seem like an abstract exercise in mathematics, but it connects directly to what we see with our own eyes. The R-S integrals are the engine that allows us to predict the intricate, beautiful patterns of light that emerge from apertures.

Of course, the exact integrals can be quite formidable to calculate. Very often, we use a further simplification called the ​​Fresnel approximation​​. This involves approximating the distance term, $r$, in the integrals, which is valid as long as our observation screen is far enough away and we don't stray too far from the central axis. We can even quantify exactly how good this approximation is by calculating a "Fresnel Validity Number" that depends on the size of our aperture, the distance, and the wavelength. This is the daily work of a physicist: knowing not only the tools but also their limits.

And the payoff for all this careful reasoning is immense. Consider the famous historical test of the wave theory of light. When Siméon Denis Poisson, a supporter of the particle theory of light, used the new wave diffraction theory to analyze the shadow of a circular disk, he calculated a shocking result: there should be a bright spot right in the very center of the shadow. He presented this as a "gotcha," a patently absurd prediction that proved the wave theory must be wrong. But when François Arago performed the experiment, the spot was there!

This phenomenon, now called the ​​Poisson-Arago spot​​, is a stunning validation of wave optics. And we can predict it perfectly with our Rayleigh-Sommerfeld tools. Using a clever mathematical trick that transforms an integral over an infinite aperture into a much simpler integral over the finite disk, the R-S theory gives the on-axis field as:

$U(0, 0, z) = A \left( e^{ikz} - e^{ik\sqrt{a^2+z^2}} \right)$

This simple formula tells a profound story. The light at the center of the shadow is the result of ​​interference​​. It's the sum of the original light wave that would have been there ($e^{ikz}$) and a new, circular wave that originates from the edge of the disk ($e^{ik\sqrt{a^2+z^2}}$). At just the right spot, these two waves interfere constructively, creating light where there "should" be darkness. This is the power and beauty of physics: a journey from intuitive ideas, through mathematical rigor and paradox, to the prediction of astonishing and observable truths about the world.

After a long journey through the mathematical thickets of wave theory, it’s easy to feel a bit lost in the symbols. We’ve painstakingly constructed a magnificent tool—the Rayleigh-Sommerfeld integrals—and it’s natural to now stand back and ask: what is it for? Is it just a formal way to calculate the fuzzy shadows behind an obstacle? The answer, I hope you’ll find, is a resounding "no!". This mathematical machinery is not the end of the story; it is the beginning. It is our ticket to a grand tour of modern science and engineering.

What we have in our hands is nothing less than the master recipe for how waves behave. And waves are everywhere. They are the light that lets us see, the sound that lets us hear, the radio signals that connect our world. By understanding this one set of principles, we unlock a staggering variety of phenomena. We are about to see how this integral, which we derived for light, becomes a universal language, spoken by physicists, astronomers, acousticians, and engineers. It's time to see the poetry this grammar can write.

Let’s begin in the familiar world of optics. At its most basic, the theory confirms our intuition and refines it. If we shine a light on an opaque screen with a hole in it, the integral tells us precisely what the light field looks like on the other side. It works for simple shapes, like a semi-circle, and allows us to calculate the intricate interference patterns that emerge on the axis, revealing a complex dance of crests and troughs where one might naively expect just a simple spot of light. Right at the screen itself, the theory gives a wonderfully simple result: the field just behind the aperture is simply the incident field multiplied by the aperture's transmission factor, a necessary "sanity check" that the theory behaves sensibly at the boundaries.

When we arrange many simple apertures together, like the slits in a diffraction grating, the Rayleigh-Sommerfeld theory elegantly morphs into the powerful language of Fourier analysis. For a regular array of slits, like a Ronchi ruling, the far-field pattern is essentially the Fourier transform of the aperture's shape. This reveals why gratings are so good at splitting light into its constituent colors: each wavelength diffracts at a slightly different angle, producing a spectrum. The theory can even tell us about the faint "subsidiary" maxima between the main bright spots, and how their weakness relative to the main peaks depends on the total number of slits in the grating. This connection between diffraction and Fourier transforms is the cornerstone of a vast field called Fourier Optics, which underlies signal processing, holography, and image filtering.

But the true magic begins when we stop thinking about just blocking light and start thinking about controlling its phase. Imagine a perfectly clear circular window. Now, suppose we place an invisibly thin piece of glass over exactly one half of it, a piece of glass engineered to delay the light passing through it by precisely half a wavelength. This corresponds to a phase shift of $\pi$ radians. What happens to the light on the central axis behind this window? Your first guess might be... something complicated. But the answer, derived directly from our integral, is astonishingly simple: it's perfect darkness. Complete cancellation. The light from one half of the aperture arrives exactly out of phase with the light from the other half, and they annihilate each other on the axis. This simple, beautiful example is a profound lesson: by subtly manipulating the phase of a wave, we can achieve dramatic, large-scale control over its intensity. This is the principle behind phase-contrast microscopy, which allows biologists to see transparent living cells, and the basis for a whole universe of "wavefront shaping" techniques.

Why stop at a simple half-circle phase shift? What if we continuously vary the phase around a circle? Suppose we create a source where the phase of the light twists like a spiral staircase, described by a term like $\exp(im\phi)$ where $\phi$ is the angle around the center. The integer $m$ is called the "topological charge" and tells us how many full twists the phase makes in one revolution. When we use our integral to see what such a beam looks like on-axis, we find another surprise: for any non-zero twist ($m \neq 0$), the intensity on the axis is again zero! The beam is hollow, shaped like a doughnut. These are not mere curiosities; they are called "optical vortices" or "twisted light," and they carry orbital angular momentum. They can be used as "optical spanners" to grab and spin microscopic particles, to encode more information into fiber optic communications, and to create super-resolution microscopes that beat the classical diffraction limit.

Another celebrity in this zoo of structured light is the Bessel beam. In an ideal world, a Bessel beam is a "diffraction-free" solution to the wave equation—its central spot doesn't spread out as it travels. Of course, in the real world, you can't create an infinitely wide beam. You must generate it through a finite aperture. What happens then? Our integral provides the answer. By modeling a perfect Bessel beam truncated by an aperture, we can calculate how the on-axis intensity, which should have been constant, now decays with distance. It shows us precisely how reality deviates from the ideal, a crucial analysis for anyone trying to use these "self-healing" beams for applications like light-sheet microscopy or precise laser cutting.

The Rayleigh-Sommerfeld integral is also an indispensable tool for the workaday world of optical design, where the enemy is not an opaque screen but the subtle imperfections of lenses and mirrors. These "aberrations" are essentially unwanted phase distortions. For example, "coma" is an aberration that makes off-axis stars look like little comets. Using the diffraction integral, we can model the wavefront distorted by coma and predict the exact shape of the blurred spot at the focus. This allows us to derive scaling laws that tell designers how the image quality degrades as the aberration gets worse, guiding the design of everything from humble camera lenses to the giant mirrors of astronomical telescopes.

It’s worth pausing to appreciate the robustness of our theoretical framework. You may recall from the previous chapter that we actually have two, slightly different Rayleigh-Sommerfeld integrals (Type I and Type II), born from different choices of a mathematical helper function. One might worry whether they give different answers to the same physical problem. They are, after all, mathematically distinct. Reassuringly, for most situations encountered in practice, especially when the observation point is many wavelengths away from the aperture, their predictions are virtually identical. One can perform detailed calculations for a specific setup, like an annular aperture, and find that the fields predicted by the two formulae are extremely close, affirming the physical consistency of the underlying theory.

So far, we have stayed in the comfortable realm of visible light. But the true beauty of a fundamental physical principle is its universality. The Rayleigh-Sommerfeld integral describes the behavior of scalar waves, and light is not the only wave in town.

Think of sound. A sound wave is a propagating wave of pressure in a medium like air or water. If a sound wave encounters an opening, it diffracts. How do we calculate the pattern? It turns out that the very same integral applies! The complex amplitude of the light field is replaced by the complex amplitude of the acoustic pressure, and the speed of light is replaced by the speed of sound. An acoustic engineer designing a concert hall, worrying about how sound from the stage diffracts around pillars, or a medical engineer designing an ultrasound transducer, is using the same fundamental wave physics as an optical engineer designing a laser system. In the high-frequency limit, the Kirchhoff approximation—a close cousin of our Rayleigh-Sommerfeld theory—predicts that a hole in a rigid barrier will transmit acoustic power with perfect efficiency, just as if the barrier weren't there. This simple result again confirms the principle's application in a new domain.

The theory's reach extends even further, into the statistical world of thermodynamics. Most light sources in nature—the sun, a flame, an old-fashioned light bulb—are not the perfectly orderly coherent waves we've been discussing. They are "incoherent," a chaotic jumble of wavelets emitted independently. Can our theory handle this mess? Yes! By extending the framework to deal with the statistical correlation between different parts of the wave (a quantity called the cross-spectral density), we can describe the propagation of partially coherent light. When this generalized theory is applied to a simple, uniform, incoherent source like a glowing hot disk, it correctly predicts how the light spreads. In fact, it perfectly reproduces a key result from the much older field of radiometry, which treats light as a simple flow of energy: the on-axis spectral density from the disk is proportional to the solid angle that the disk subtends from the observation point. It's a beautiful moment of synthesis, where the detailed microscopic wave theory, when averaged appropriately, gives rise to the simple, macroscopic laws of energy transport.

And now for the grandest stage of all: the cosmos. We built our theory on the foundation of a flat, Euclidean space. But Einstein taught us that in the presence of massive objects, spacetime itself is curved. What happens to a wave propagating near, say, a black hole? The fabric of space and time is warped, and this warping affects the path of the light. In the language of optics, it's as if the vacuum itself has acquired a spatially varying refractive index. This is the phenomenon of gravitational lensing. In a remarkable display of theoretical power, we can take our humble diffraction integral and adapt it to this extraordinary environment. In a thought experiment, we can modify the path length term in the integral to include the extra delay caused by the gravitational field of a Schwarzschild black hole. By doing this, we can predict how a diffraction pattern from an aperture would be altered by the presence of the black hole. While this is a purely theoretical calculation performed under several simplifying assumptions, it is a breathtaking demonstration of the adaptability of the wave concept. It shows how the same fundamental idea—summing up all the possible wavelets, each with its own phase determined by its path—can be transplanted from a lab bench to the edge of an event horizon.

From the hum of an auditorium to the twisting of light in a microscope and the bending of starlight around a black hole, the same song is being sung. The Rayleigh-Sommerfeld integrals provide the sheet music, revealing a deep and unexpected unity in the behavior of waves across vast, disparate scales of the physical world.