Squeezed States of Light

In the quantum realm, even a perfect vacuum seethes with inherent fluctuations, imposing a fundamental noise floor on all measurements known as the Standard Quantum Limit (SQL). For decades, this limit seemed to be an absolute barrier to precision. Squeezed states of light represent a profound and elegant solution to this challenge, offering a way to manipulate quantum uncertainty itself to see beyond the noise. By engineering light to be quieter in one property at the calculated expense of another, we can perform measurements once thought impossible.

This article provides a comprehensive overview of this non-classical resource, bridging its fundamental principles with its revolutionary impact. It addresses the knowledge gap created by the SQL by explaining how we can strategically circumvent it. The reader will discover the theory, creation, and transformative use cases of squeezed light. The first chapter, ​​Principles and Mechanisms​​, demystifies the quantum physics of squeezing, from the Heisenberg Uncertainty Principle to the nonlinear optical processes that generate it. Subsequently, the chapter on ​​Applications and Interdisciplinary Connections​​ showcases how this technology is currently pushing the boundaries of gravitational wave detection, precision sensing, and quantum computing, illustrating the power of controlling the quantum vacuum.

Imagine a perfectly still pond. The water, to your eye, is flat and motionless. But if you could look with a powerful enough microscope, you would see a constant, restless dance of water molecules. The quantum world is much the same. Even a perfect vacuum, the darkest, emptiest space you can imagine, is not truly empty. It is a roiling sea of "virtual particles" and fluctuating fields. This inherent, unavoidable quantum jitter is what we call ​​vacuum fluctuations​​.

For a single mode of light—think of it as a single note played on a cosmic guitar string—this vacuum is its ground state, its lowest possible energy. We can model this mode of light as a ​​quantum harmonic oscillator​​. Its properties can be described by quantum operators that are analogous to the position $x$ and momentum $p$ of a pendulum. For light, we call these the ​​quadrature operators​​. We can label them $X$ and $P$. You can picture them as the amplitude and phase of the light wave, respectively, plotted on a 2D "phase space" map.

Even in the vacuum state $|0\rangle$, where there are no photons on average, these quadratures are not perfectly zero. They fluctuate. The strange and beautiful rule of quantum mechanics, the ​​Heisenberg Uncertainty Principle​​, tells us that we cannot know both of these values with perfect precision simultaneously. The more precisely we know the "position" $X$, the less precisely we know the "momentum" $P$. Mathematically, the product of their uncertainties (standard deviations, $\Delta$) must be at least a certain minimum value:

(Here we use natural units where a fundamental constant, $\hbar$, is set to 1). For the vacuum state, and also for the intense, highly stable light from a laser (called a ​​coherent state​​), nature is as fair as can be. The uncertainty is minimal, and it is distributed equally between the two quadratures. The "fuzziness" of our knowledge in phase space forms a perfect circle. This fundamental noise floor, where the noise is equally shared, is known as the ​​Standard Quantum Limit (SQL)​​. For decades, it was thought to be an unbreakable barrier for measurement precision.

But what if your experiment only depends on one of these values? What if, to detect a faint gravitational wave, you only need to measure the light's amplitude, $X$, with breathtaking precision? Do you still have to pay the price of uncertainty in $X$ just because $P$ is also uncertain?

Here lies one of the most ingenious tricks of modern quantum physics. The answer is no! The uncertainty principle sets a limit on the area of the uncertainty region in our phase space map, not its shape. While we cannot shrink the area below the minimum, we can deform it. We can squeeze the circle into an ellipse.

This is the central idea of a ​​squeezed state of light​​. We can reduce the uncertainty in one quadrature, say $\Delta X$, making it smaller than the standard quantum limit ($\Delta X \frac{1}{\sqrt{2}}$). This is called ​​squeezing​​. But there's no free lunch in physics. To preserve the total area of uncertainty, the ellipse must stretch in the other direction, increasing the uncertainty in the other quadrature, $P$, making it larger than the SQL ($\Delta P > \frac{1}{\sqrt{2}}$).

This is not just a theoretical fantasy. We can quantify this effect precisely. As shown in the thought experiment of problem, by applying a "squeezing" process with strength $r$, the uncertainties in the two quadratures become:

The ratio of the stretched noise to the squeezed noise is a startling $\exp(2r)$! In a similar analogy with a mechanical oscillator, the position uncertainty might be squeezed at the cost of momentum uncertainty. For any squeezing $r > 0$, we have beaten the standard quantum limit for one variable. The product of the uncertainties remains at the minimum allowed value, $(\frac{1}{\sqrt{2}} e^{-r}) \cdot (\frac{1}{\sqrt{2}} e^{r}) = \frac{1}{2}$, so we are still playing by Heisenberg's rules. We've just learned to be clever about them.

How do we perform this magical-seeming operation? In the language of quantum mechanics, we do it by applying a ​​squeezing operator​​, denoted $S(\xi)$, to an initial state, typically the vacuum. This operator has a very specific mathematical form:

Here, $a$ and $a^\dagger$ are the fundamental ​​annihilation and creation operators​​. They are the tools we use to destroy and create single photons. Notice the peculiar terms in the operator: $a^2$ and $(a^\dagger)^2$. The operator doesn't work on one photon at a time; it annihilates photons in pairs ($a^2$) or creates them in pairs ($(a^\dagger)^2$). This is the signature of squeezing.

This pairing has a profound consequence. If we apply the squeezer to the vacuum state, what do we get? We get a state that is a quantum superposition of only even numbers of photons! As calculated in, the squeezed vacuum state $| \xi \rangle$ looks like:

where the $C$'s are complex coefficients. It's a state with zero photons, plus a bit of a two-photon state, plus a smaller bit of a four-photon state, and so on. This is radically different from the coherent state of a laser, which contains a mix of all possible photon numbers. The fact that the photons are born in pairs is a deep indication of the non-classical nature of squeezed light.

The action of this operator can also be viewed from a different angle. Instead of changing the state, we can see how it transforms the fundamental operators themselves. This transformation, known as a ​​Bogoliubov transformation​​, mixes the creation and annihilation operators. After the transformation, the "new" annihilation operator, $b = S^\dagger a S$, becomes a linear combination of the old $a$ and $a^\dagger$. It is this mixing of particle creation and annihilation that effectively deforms the geometry of quantum phase space, turning our uncertainty circle into an ellipse. For this whole procedure to be physically consistent, the transformation must preserve the fundamental commutation rules of quantum mechanics. This imposes a beautiful constraint on the coefficients, $|u|^2 - |v|^2 = 1$, ensuring our new description still lives in a valid quantum world.

The squeezing parameter, $\xi = r e^{i\theta}$, is a complex number. We've seen that its magnitude, $r$, determines how much we squeeze. Its phase angle, $\theta$, is just as important: it determines the ​​squeezing angle​​, or the orientation of the uncertainty ellipse in phase space.

This means we can't just measure any old quadrature and expect to see reduced noise. We have to "look" at the light from the right angle—an angle that aligns our measurement with the squeezed, narrow axis of the ellipse. Imagine trying to measure the width of an ellipse with a ruler. If your ruler is aligned along the long axis, you'll measure a large value. If it's aligned along the short axis, you'll measure a small one.

In an experiment, we can choose our "measurement angle," let's call it $\vartheta$, by adjusting the phase of a reference laser beam. A generalized quadrature is then defined as $X(\vartheta)$. The variance, or noise, we measure in this quadrature depends on the relationship between our measurement angle $\vartheta$ and the squeezing angle $\theta$. As derived in the context of problem, the variance has a term that looks like $-\cos(2\vartheta - \theta)$.

The crucial insight, formalized in, is that to see the minimum possible noise, we must be clever and choose our measurement angle $\vartheta$ such that it aligns perfectly with the squeezed axis (when $2\vartheta = \theta$). When we do this, the cosine term becomes 1, and the noise is maximally suppressed. The minimum achievable variance is not zero, but a new floor set by the squeezing strength:

This is the ultimate payoff. By preparing a squeezed state and measuring along its quiet axis, we can perform measurements with a precision that was once thought to be fundamentally impossible.

We've triumphed over the noise in one quadrature, but we know there's no free lunch. We stretched the noise in the other quadrature. But there's another, more subtle trade-off at play: the relationship between phase (related to the quadratures) and photon number.

What happens to our knowledge of the number of photons in the beam when we create a quadrature-squeezed state? The uncertainty in the photon number, $\Delta N$, explodes! The very process of creating photons in pairs, which squeezes the quadrature noise, makes the number of photons arriving at any given time extremely uncertain. For a squeezed vacuum state, the photons arrive in "bunches" or "clumps," a behavior known as ​​super-Poissonian​​. The stronger the squeezing, the "clumpier" the light becomes.

It is fascinating to note that one can play the game the other way around. It's possible to create "number-squeezed" states where the photon number uncertainty is reduced below the standard (Poissonian) limit. Such a state has a Fano factor $F$ (the ratio of variance to mean) less than 1. As explored in, this number-squeezed light exhibits a behavior called ​​anti-bunching​​. The photons in such a beam are more evenly spaced than in a random stream, arriving one by one in an orderly fashion. This highlights the rich and varied textures that light can assume in the quantum realm.

So, how do we actually build a device that creates these exotic states? The key lies in the field of ​​nonlinear optics​​. Most materials, like a window pane, are linear: the properties of the material don't change based on the intensity of the light passing through it. But some special crystals, when hit with a very intense laser beam, behave differently. The light itself is strong enough to alter the optical properties of the material, causing photons to interact with one another in strange ways.

One such process is used in a ​​Degenerate Parametric Amplifier (DPA)​​. In a DPA, a strong "pump" laser beam of frequency $2\omega$ is shone on a nonlinear crystal. The intense field stimulates the crystal to convert single, high-energy pump photons into pairs of identical, lower-energy "signal" photons, both with frequency $\omega$. A single photon at $2\omega$ becomes two photons at $\omega$.

This physical process, the creation of photon pairs, is the direct realization of the $(a^\dagger)^2$ term in our abstract squeezing operator! The Hamiltonian (the operator for energy) that describes this interaction is, after some simplification, exactly the kind of operator we need:

where $\chi$ represents the strength of the nonlinear interaction. When we let an initial vacuum state evolve under this Hamiltonian for a time $T$, the output is a squeezed vacuum state. The squeezing parameter $r$ is simply given by the product of the interaction strength and time, $r = \chi T$. The average number of photons produced, $\langle N \rangle = \sinh^2(\chi T)$, perfectly matches the abstract predictions. Here, in a humble crystal on a laboratory bench, the elegant mathematics of Bogoliubov transformations and squeezing operators is made manifest, forging a new kind of light to help us probe the deepest secrets of the cosmos.

Having journeyed through the strange and beautiful principles of squeezed light, we might naturally ask: What is it good for? Is it merely a curiosity of the quantum world, a solution in search of a problem? The answer, it turns out, is a resounding no. The ability to tame the wild uncertainty of the quantum vacuum is not just a neat trick; it is a revolutionary tool that is pushing the boundaries of science and technology. We are beginning to build instruments that see farther, measure more precisely, and compute in ways that were once the stuff of science fiction, all by being clever about how we handle nothingness.

Let’s embark on a tour of these applications, from the colossal detectors listening for the whispers of spacetime to the delicate dance of atoms and the very architecture of future computers. You’ll see that the principle is always the same: by quieting nature’s inherent noise in one respect, we can perceive a reality that was previously hidden in the quantum fog.

Perhaps the most spectacular application of squeezed light is in the search for gravitational waves. Imagine trying to hear a pin drop in the middle of a thunderstorm. That’s the challenge faced by instruments like the Laser Interferometer Gravitational-Wave Observatory (LIGO). The “pin drop” is a minuscule distortion of spacetime itself, caused by cataclysmic events like the merging of two black holes billions of light-years away. The “thunderstorm” is a cacophony of noise, from seismic vibrations to the thermal jiggling of atoms in the mirrors.

But even if we could eliminate all those classical noise sources, we would hit a fundamental wall: the quantum noise of light itself. The powerful lasers in LIGO’s arms are composed of photons, and the random arrival of these photons at the detector—shot noise—creates a fluctuating pressure on the mirrors, obscuring the faint signal of a passing gravitational wave. This isn't a technical flaw; it's the very graininess of the quantum world.

This is where squeezed light performs its magic. By replacing the ordinary vacuum that enters the interferometer's unused port with a specially prepared squeezed vacuum, physicists can reduce the quantum noise in the very measurement they need to make. Of course, the universe exacts its price; the noise in the other, unmeasured quadrature increases, but we can simply choose to ignore it! This technique has allowed LIGO and its partner observatories to significantly increase their sensitivity, effectively letting them see deeper into the universe. The amount of noise reduction is often spoken of in decibels (dB), a direct measure of the squeezing parameter $r$ that we have discussed. A few dB of squeezing can mean the difference between seeing a few events per year and dozens.

But this is not a simple plug-and-play upgrade. The real world is messy. The precious squeezed state must travel through optical components, like beam splitters and mirrors, which are never perfectly lossless. Every photon that gets lost is replaced by a random photon from the environmental vacuum, which is noisy and un-squeezed. This process degrades the squeezing, mixing the quiet order back with chaos. A beam splitter with transmissivity $T$, reflectivity $R$, and loss $L$ will turn a perfectly squeezed noise level of $e^{-2r}$ into a diluted version, roughly $T e^{-2r} + R + L$, where the $R$ and $L$ terms represent the intrusion of vacuum noise. Furthermore, the electronics in the detectors have their own noise, which adds to the din. To achieve a target level of noise suppression, engineers must start with a much higher degree of squeezing to overcome these inevitable losses and imperfections. Even the optical cavities used to build up the light power can twist and rotate the squeezing effect in a frequency-dependent way, a subtle dance that must be carefully choreographed. The story of squeezed light in LIGO is a powerful lesson in both the beauty of a quantum idea and the grit of real-world engineering.

The principles that enable LIGO to hear the cosmos are not limited to astrophysics. They apply to any optical measurement that is limited by shot noise. Consider a high-precision sensor designed to detect a faint chemical trace or a tiny displacement. If its sensitivity is limited by the statistical patter of photons from its laser source, replacing that laser with a squeezed light source can directly improve its signal-to-noise ratio (SNR). The degree of improvement depends critically on how good your detector is. A perfect detector with 100% quantum efficiency would reap the full benefit of squeezing, while a lossy detector would see a more modest gain, as it inadvertently mixes in vacuum noise. This principle is now being explored for everything from microscopy to navigation.

This quest for precision also leads us to ask a more profound question. Is there an ultimate limit to how well we can measure something? Quantum mechanics says yes. For a fixed number of "probes" (say, photons), there is a "Heisenberg Limit" on measurement precision, a bound far more stringent than the standard shot-noise limit. The exciting discovery is that squeezed light is a key ingredient to approaching this ultimate boundary. By cleverly combining a strong, classical-like coherent state with a faint squeezed vacuum state in an interferometer, we can create a probe that is exquisitely sensitive to tiny phase shifts. The Quantum Fisher Information, a measure of the maximum possible information one can extract, reveals that a combination of high laser power (from the coherent state) and strong squeezing can lead to sensitivities that scale far better than what's possible with classical light alone. We are not just making better measurements; we are pushing against the fundamental limits of knowledge.

So far, we have viewed squeezed light as a tool for passive observation. But what if we used it as a raw material for construction? This is precisely the role it plays in the burgeoning field of quantum information and computation.

One of the essential resources for quantum computing is entanglement, the mysterious connection between two or more quantum particles that Albert Einstein called "spooky action at a distance." It turns out that a recipe for creating a powerful form of entanglement is beautifully simple. Take a beam of light squeezed in its amplitude quadrature. Take a second beam squeezed in its phase quadrature. Mix them on a simple 50:50 beam splitter. What comes out is no longer two separate beams, but a single, unified two-mode state where the output beams are deeply entangled. The amount of squeezing you put in, quantified by the parameter $r$, directly determines the amount of entanglement you get out. With a tabletop optical setup, one can generate the very "spooky action" that is the fuel for quantum algorithms.

This is just the beginning. By taking many squeezed light sources and weaving them together with a network of beam splitters, scientists are building "cluster states." These are massive, entangled webs of light that serve as the substrate for measurement-based quantum computing. In this scheme, the computation is performed not by applying a sequence of logic gates, but by making a series of measurements on the pre-prepared cluster state. The quality of this state is everything. Here, we encounter another one of nature's fascinating trade-offs. The state is judged by "nullifiers"—operators whose measurement outcome should be zero for a perfect state. Any deviation from zero represents an error. The variance of these nullifiers has two main sources of error: the finite amount of initial squeezing, and imperfections in the entangling operations. The finite squeezing error gets smaller as the squeezing parameter $r$ gets larger, which makes sense. But the error from imperfect gates is multiplied by the anti-squeezing factor, $e^{2r}$! This means that as you squeeze harder to reduce one type of error, you make the system dramatically more sensitive to the other type. It is a delicate balancing act, a profound lesson in the two-faced nature of squeezing and anti-squeezing.

The influence of squeezed light extends into an ever-widening circle of scientific disciplines. In atomic physics, laser cooling is a workhorse technique used to chill atoms to temperatures billionths of a degree above absolute zero. This is done by bombarding the atoms with laser photons that slow them down. However, the random kicks from photon absorption and re-emission cause the atom's momentum to diffuse, creating a "heating" effect that sets a fundamental limit on the final temperature. What if the photons were not so random? Using squeezed light, which has sub-Poissonian photon number statistics, can actually reduce this momentum diffusion. The more "orderly" arrival of photons imparts fewer random kicks, potentially allowing atoms to be cooled to even lower temperatures than possible with classical light.

Finally, let’s turn the story on its head. We have discussed how to generate and use squeezed light. But could nature be generating it without our help? Recent theoretical work suggests that squeezed states might arise in the most unexpected places. For example, in a type of advanced microscope that uses a tiny, vibrating metal tip to see details at the nanoscale, the rapid mechanical motion of the tip can parametrically modulate the local electromagnetic vacuum. This process, in theory, is precisely the mechanism of a squeezing operation. The very act of a vibrating nanostructure could be "squeezing" the vacuum around it, generating non-classical light.

From the vastness of intergalactic space to the motion of a single atom and the tip of a nanoscale probe, squeezed light is a unifying concept. It represents a new level of control over the quantum world, and with this control comes an exhilarating power to discover, to measure, and to build. The journey into the quiet side of light has only just begun.