 This video, more than most that I've made, starts off simple and then ramps up quickly, so as always I encourage you to skip ahead or rewind as necessary. The way that electrons orbit an atom is frequently compared to the way that planets orbit the Sun. It's worthwhile briefly looking at the similarities and differences between these two structures. With the Solar System, a planet or artificial satellite can orbit arbitrarily close to or far away from the Sun. The closer an object orbits the Sun, the more tightly it is held by gravity, the less energy it has. To go from a small orbit, out to a further one, energy must be expended, just as you would have to perform work against gravity to lift an object from the surface of the Earth. So in short, objects can orbit the Sun at any distance, but the further out they are, the more total energy they have. Electrons orbiting an atomic nucleus are different in that they are governed by quantum mechanics. An electron can take one of a number of specific or quantized orbits, more accurately called quantum states. For an electron in a given state, the position is no longer defined with certainty, but rather is probabilistic in nature. The energy, however, is fixed. Just as with the planets, the states where the electrons orbit closer to the nucleus on average, have lower energies. So in short, electron energies can take only precise, quantized values and nothing in between. This picture is further complicated by the fact that whereas the planets exert only a weak force on one another, so for example the gravitational pull of Jupiter on Earth is quite small and that of Mercury is downright negligible, the repulsive force between multiple electrons in an atom is quite strong. This means that adding an additional electron to an atom changes the energy levels because of their mutual repulsion and also makes it much harder to calculate them compared to just a single electron orbiting with no repulsion. The long and short of all this is that an electron and an atom can only take one of a set of fixed energy levels. The lowest possible energy is called the ground state and above that are excited states. An electron can be excited up in energy if another atom or free electron collides with it. The colliding particle loses energy which the electron gains. The electron can also be excited by absorbing a photon, but this photon must be carrying almost exactly the same energy as the gap between the levels in question. All of these processes are reversible, so a free electron may come in, de-excite a bound electron and carry off the excess energy. If an excited bound electron is left alone for long enough, it will decay after a certain time, which is random. By carefully managing these atomic processes, it is possible to generate useful light. In the United Kingdom, for example, we have a lot of street lamps where an electric current, in other words the movement of energetic electrons, is passed through sodium vapor. The moving free electrons will collide and excite the bound electrons in the sodium atoms. The excited electrons then decay, emitting photons of yellow light at 589 nanometers in wavelength and an energy corresponding exactly to the difference between the ground and excited states. But this is not a laser. Each of the atoms emits a photon at random times and moving in random directions, so this does not produce the kind of tight beam that is characteristic of a laser. Normally a sodium light will have a reflector behind it, so the light goes broadly in one direction, but it still spreads out into a wide area. The phenomenon we have seen so far is more precisely called spontaneous emission, where an excited electron left alone for some amount of time will emit a photon in isolation. But if there is already a photon with the right energy passing by the atom, it might cause stimulated emission of a new one. The oscillating electric field of the original photon interacts with the electron leading to the emission. Because of this, stimulated emission is no longer random. The two photons are synchronized in time and moving in the same direction. Also, like all quantum processes, stimulated emission is not guaranteed but merely has a probability to happen. The energy of the photon has to match the gap between energy levels very closely, which typically means you have to have atoms of the same element for this to work. The energy levels of sodium are different to neon and so on. Anyway, at the end of all that, you just need to understand that an excited electron will emit a photon by spontaneous emission if it's left alone, or by stimulated emission if an existing photon comes in. These processes are harnessed with a laser to create a chain reaction. One photon emitted spontaneously stimulates the emission of a second, two soon turn to four, then eight, and so on. The photons are all moving at the same time in the same direction. The growth in strength of the laser light is exponential, where the speed of the growth is called the gain. The crucial ingredient to have gain is to have plenty of atoms with excited electrons, and I'll come on to how they're excited in a moment. But in a laser, these atoms are assembled into a material referred to as the gain medium. The gain medium has to have atoms of the same type, but they can be in a solid, liquid, gaseous, or even plasma state. The atoms may, however, be mixed in with other materials, so for example in a moment we will look at a type of laser where individual neodymium atoms, usually metallic, are embedded in glass. So a gain medium will send out bunches of photons. After all, laser is an acronym for light amplification by stimulated emission of radiation. This is good, but we are back to the problem of the sodium light. The initial spontaneous emission is oriented randomly, so these groups of photons are going to be leaving the gain medium, travelling in all the different directions. However, there is a way around this. The gain medium is made into a long but narrow cylinder. If a photon is spontaneously emitted travelling in the short dimension of the cylinder, it will double itself only a few times and then quickly leave the gain medium. If it's travelling the long way through the cylinder, it will double itself for very many times and form a strong beam when it finally leaves. The difference between these two is literally exponential. Another way to make this effect even stronger is to put mirrors at either end of the cylinder, so the beam will go back and forth through the gain medium many times and grow even larger. In order to actually be useful, one of the mirrors is partially reflective and partially transparent. Some of the light will then leave the laser for its intended purpose. All the photons are travelling in largely the same direction in a tight beam and each one carries almost the same energy, meaning they all have the same colour. As we have seen, the laser only creates more photons at the expense of excited electrons in atoms. Eventually, the gain medium will run out of these atoms and that initial exponential growth will taper off. More atoms must be continuously energised into an excited state in a process called pumping. Eventually, as the strength of the laser beam grows, it will reach an equilibrium. Atoms will be pumped into the excited state at a rate of r atoms per second. New photons will be emitted from the gain medium at a rate of r per second and photons will leave through the output mirror at a rate of r per second. This is usually referred to as a continuous wave laser and it will work steadily as long as energy is provided for the pumping. There is a way to create a much more powerful but short laser pulse with a technique called Q-switching. Imagine removing the mirror at one end of the laser. Now photons can no longer make continuous round trips through the gain medium to build up their number. Because there is no longer a powerful beam, there is nothing to deplete those excited atoms and so more and more of them become pumped up to become excited. If the mirror is suddenly put back, a beam will quickly grow to much larger intensity than before because there are now so many more excited atoms. You're quite literally charging your laser above the level it would normally be at and then releasing that energy in a big burst. It will of course quickly deplete the number of excited atoms so the pulse will not last very long. The only caveat is that the mirror has to be put back quickly, perhaps on the order of microseconds, much faster than it can be physically moved, much faster also than a shutter can open. Instead there are electronic methods such as applying a voltage to make crystals transparent and so on. The technique is called Q-switching because the quality factor of the mirror, how well it reflects light, has to be quickly switched from low to high. So back to pumping. How do you make the electrons in atoms excited to the right level? And as a shorthand I will just say that the whole atom is excited rather than the electron. Firstly, let me clarify that when I said that an excited atom can emit a photon, it doesn't necessarily have to drop down all the way to the ground or lowest energy level. In fact, for reasons I'll explain in a moment, most don't. All that is needed is that an atom starts in a high energy state called the upper laser level and then drops down an energy to a lower laser level. But it doesn't necessarily need to be the lowest possible. As we've seen, it's possible to excite atoms with an electric current or with incoming light. For the latter, it's possible to use an ordinary light source, like a lamp, which also takes electrical input directly. The problem with the laser is that it's not enough to just have some atoms in the upper level. Remember what I said about quantum processes, they're reversible. When photons encounter an atom in the upper level, stimulated emission will create a new photon. When they encounter an atom in the lower level, a photon will be lost by stimulated absorption, the inverse process. These normally try to minimize their energy so ordinarily there are more atoms in the lower energy levels than in the upper ones. A laser beam will then lose more photons by absorption than it gets back by emission. There is no longer gain. The pumping needs to flip this situation, attaining what is called population inversion. More atoms in the upper level than the lower. Okay, let's do that. Let's excite up more than half the atoms. The inverse processes are going to hit us hard once again. Let's say we put a bunch of photons from a lamp to try and excite our gain medium. Once the fraction of atoms in the upper level gets above 50%, the incoming light is going to start causing more losses by stimulated emission than they put back. It's the opposite problem of needing population inversion. The way to solve this problem is to make use of more than just two atomic energy levels. Let's look at an example system where the ground state is the lower laser level, above that an energy is the upper laser level, and there is a third level a little higher in energy still. As usual I will pull some numbers out of my behind purely for illustrative purposes. 40% of atoms are in the lower level and 10% in the highest level. This is perfectly fine to pump. Once they're pumped to the third level, atoms decay down to the upper lazing level, giving up the excess energy as waste heat. This is also fine. Because this loss of energy happens so quickly, atoms don't stay on the third level, but a large number ends up in the upper laser level, 50% of the total. Now for the lazing. With a 50-40 split, there are more atoms in the upper level, therefore there is a population inversion and gain. The laser works. The downside of such a three level laser is that a total of 60% of the atoms must be kept excited above the ground state. That requires a lot of power. More efficient is a four level laser. The two laser levels are both above the ground state, and there is a fourth level above them again. Get ready for some more butt numbers. 95% in the ground state and 1% in level four. Pumping is fine. Quick decay down to level three, the upper laser level, leaving 3% of all atoms in that state. Many atoms in the lower laser level decay down to the ground state quickly, so only 1% are in this level. As a result, there is still a population inversion and therefore gain, but only 5% of the atoms are in an excited state at any one time, so the energy required to run this four level laser is lower than the three level laser. With all that in mind, let's look at some common and some not so common laser designs. The ruby laser was the first type to ever operate. The gain medium is a classic crystal structure associated with lasers, a cylindrical ruby rod with the same chemical composition as the gemstone, much longer than it is wide. Ruby is mostly made of aluminum oxide with some chromium mixed in. These atoms give it the red color and are the ones which produce the laser light. Ruby lasers might typically be pumped by a xenon flash lamp. A very high current is passed through xenon gas, actually it becomes a plasma, and it emits light which excites the chromium. Usually there are mirrors all around the flash lamp to direct the light into the ruby. It is a quintessential three level laser. A similar design with neodymium has been used to make the most powerful lasers in existence, such as those at the National Ignition Facility. Neodymium atoms, or actually ions as they are each missing three electrons, are embedded in a yttrium aluminum garnet crystal, or simply in glass. Again, there are different ways to pump it, but the key difference is that this is a four level laser and therefore more efficient. Something which is not so technologically useful but which you will probably encounter as a teaching tool if you ever take a proper course on lasers is the helium neon laser. Both these noble gases are mixed together. An electric current excites the helium to a particular level. What's interesting is that this level matches a particular energy level in neon, which happens to be the fourth level in a four level laser system. When excited helium collides with neon, it gives up its energy and effectively pumps the neon. There are other lasers with a gaseous gain medium like the CO2 laser as well. By far the most common type of laser in the world, present in barcode scanners and in CD drives when they used to be a thing, is the semiconductor or diode laser. Semiconductors are a lot more complicated than even the quantum picture of a single atom I talked about earlier. Rather than specific levels, there are energy bands, so the laser operates between those instead. Usually a semiconductor like silicon has a small amount of impurity atoms added or doped, which have either too many or too few electrons. An n-type material is one with too many electrons which are then free to move. A p-type material is one with too few electrons, leading to so-called holes left behind. The holes act and move like positive particles. If a neighboring electron moves left into the hole, a new one is left behind as if the old one had simply moved right. An ordinary diode is made by sandwiching p-type and n-type materials next to each other. Two regions meet at a junction where the electrons and holes recombine to cancel out. This all works a bit like transitions between levels and atoms. They can recombine spontaneously or be stimulated by a photon. Normally a photon would have to give up its energy by stimulated absorption to cause this recombination. However, applying voltage across the junction makes it energetically favorable for an electron and hole to recombine, so then existing photons can stimulate the emissions of new ones. Feeding electrical energy in creates gain and therefore lasing. Obviously I've simplified a lot here and the diode laser requires a lot more engineering than I've discussed, but this type of laser can now be built for pennies. The design is simple in that partial reflection happens from the output faces of the semiconductor crystal itself. Huge arrays of semiconductor lasers can then be used to pump other, more powerful types of laser. Now for some more unconventional laser types. Light in the electromagnetic spectrum beyond ultraviolet is strongly ionizing, which means it would be absorbed by any material including air until it has turned that material into a plasma. That's not to say that there aren't any reasons to create a laser in the far ultraviolet or x-ray part of the spectrum, but it would have to exist in space or a vacuum chamber. The gain medium must also be a plasma, or it would soon become one. The gain medium for this type of laser can be pumped by other lasers in the visible part of the spectrum by intense electric currents or even nuclear explosions. The idea of Ronald Reagan's Star Wars program was to detonate a nuclear warhead in space which would obviously release a lot of energy. This energy would somehow be channeled to pump a laser system creating an x-ray beam to shoot down enemy missiles. A more controlled way to create a laser with short wavelengths is called a capillary discharge. A long cylindrical capillary has a noble gas being pumped through it. A strong electrical current is briefly discharged through the gas, quickly ionizing it and exciting those remaining electrons that are still in the ions. At this point, there are no mirrors at either end of the gain medium. Many photons travelling the long way down the capillary get amplified extremely quickly and in a matter of nanoseconds the pulse is over. Another proven technique to make laser light at any wavelength is the free electron laser. When very fast moving electrons accelerate, they emit radiation. This is actually a problem for accelerators like the LHC because all charged particles lose energy in this way. Why not harness this radiation? A beam of electrons is generated by a linear accelerator. This beam is then passed through a region of alternating magnets so that the magnetic force causes a side-to-side acceleration and hence a tight forward beam of photons. The initial photons effectively stimulate further emission if the wavelength matches the spacing of the magnets and the electrons bunch themselves up with the same spacing. The output of such a free electron laser also grows exponentially up to a point. The neat thing is that the wavelength of the photons depends on the energy of the electrons coming out of the linear accelerator. Not only can the latter be designed with a huge range of energies in mind, but those energies can be varied on the fly. A free electron laser can be built to work with anything from microwaves to x-rays and then dial in the precise wavelength of photons desired within a certain range. This is useful for experiments and even weapons. The color of the beam can be rapidly varied to the best atmospheric conditions. There is also a specialist technique called lasing without inversion. If a laser has a very large energy gap between its lower and upper levels, this would mean that it would take a lot of pump energy to create a population inversion, especially if it's not possible to have a four-level laser. The gain medium needs to only be pumped by a small amount, not enough to create a population inversion. Then, a relatively small amount of energy can be injected to the atoms in the lower laser level, not to take them all the way up, but to transfer some of them to a low-lying energy level not involved in the lasing. This is the best way I can explain it in brief, but if you've done the equivalent of a university-level quantum mechanics course, here is a paper with all the details which should be understandable. So to sum up, lasers make use of the fact that energy is released as light, as photons, and an electron orbiting a nucleus goes from one orbit high up in energy to one lower down. Photons stimulate the emission of more photons of exactly the same wavelength or color moving in the same direction. A laser is physically engineered to put out a tight beam of light by carefully selecting its shape and adding mirrors is required. It can be continuous or pulsed. A given laser uses one type of atom to actually create the light, but these atoms may be on their own in a gas or a plasma, embedded in a crystal or some other arrangement. And in some cases, like the free electron laser, atoms are not even required. Thank you for watching.