MISN-0-246 WAVE-PARTICLE DUALITY: LIGHT by E. H. Carlson 1. The Problem Posed by Light a. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 b. Classical Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 c. Classical Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 d.

The Nature of Light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2. Both Aspects in One Experiment a. The Single Slit Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 b. Particle Model Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 c. Wave Model Predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 d. Both Aspects Observed Simultaneously . . . . . . . . . . . . . . . . . . 6

WAVE-PARTICLE DUALITY: LIGHT

c c

PM

PM

3. The Quantum Field Theory of Light a. Light Is Not Particles, Not Waves . . . . . . . . . . . . . . . . . . . . . . . 7 b. Quantum Field Theorys Description . . . . . . . . . . . . . . . . . . . . 7 c. Classical EM Wave Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 4. Field Theory Applications a. Strange One- and Two-slit Predictions . . . . . . . . . . . . . . . . . . . 9 b. Strange Spherical Wave Predictions . . . . . . . . . . . . . . . . . . . 9 c. Designing and Using an Astronomical Telescope . . . . . . . . 10 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Project PHYSNET Physics Bldg. Michigan State University East Lansing, MI

ID Sheet: MISN-0-246 THIS IS A DEVELOPMENTAL-STAGE PUBLICATION OF PROJECT PHYSNET The goal of our project is to assist a network of educators and scientists in transferring physics from one person to another. We support manuscript processing and distribution, along with communication and information systems. We also work with employers to identify basic scientic skills as well as physics topics that are needed in science and technology. A number of our publications are aimed at assisting users in acquiring such skills. Our publications are designed: (i) to be updated quickly in response to eld tests and new scientic developments; (ii) to be used in both classroom and professional settings; (iii) to show the prerequisite dependencies existing among the various chunks of physics knowledge and skill, as a guide both to mental organization and to use of the materials; and (iv) to be adapted quickly to specic user needs ranging from single-skill instruction to complete custom textbooks. New authors, reviewers and eld testers are welcome. PROJECT STAFF Andrew Schnepp Eugene Kales Peter Signell Webmaster Graphics Project Director

Title: Wave-Particle Duality: Light Author: E. H. Carlson, Michigan State University Version: 2/1/2000 Length: 1 hr; 20 pages Input Skills: 1. Vocabulary: particle trajectory, Newtons second law (MISN-015) or (MISN-0-409); photon energy, momentum, wavelength, frequency (MISN-0-212); two-source interference, diraction (MISN0-235) or (MISN-0-432); Snells law of Refraction, polarization of light (MISN-0-222); electrostatic force (MISN-0-115) or (MISN-0419); Maxwells Equations (MISN-0-132) or (MISN-0-429). 2. Apply the photon energy-momentum-frequency relations (MISN0-212). Output Skills (Knowledge): K1. Vocabulary: quantum eld theory, wave-particle dualism. K2. Contrast the quantum eld theory description of light with that of the classical particle and wave models. Describe the predictive power of each of these. Output Skills (Rule Application): R1. Give reasons why any given light phenomenon can or cannot be satisfactorily described as each of these: a. classical particle b. classical EM wave c. quantum eld Evaluation: Stage 1

ADVISORY COMMITTEE D. Alan Bromley E. Leonard Jossem A. A. Strassenburg Yale University The Ohio State University S. U. N. Y., Stony Brook

Views expressed in a module are those of the module author(s) and are not necessarily those of other project participants. c 2001, Peter Signell for Project PHYSNET, Physics-Astronomy Bldg., Mich. State Univ., E. Lansing, MI 48824; (517) 355-3784. For our liberal use policies see: [Link]

MISN-0-246

MISN-0-246

WAVE-PARTICLE DUALITY: LIGHT by E. H. Carlson 1. The Problem Posed by Light

1a. Overview. In classical physics, especially mechanics and electricity and magnetism, two distinct and very useful concepts have been invented and used repeatedly, that of particles and that of waves. However, to explain phenomena on the atomic level of size, the distinction between waves and particles becomes blurred, and an important modication, involving probability, must be made. Because the separate concepts of particle and of wave are so natural and satisfying for phenomena on the human scale it is somewhat dicult to explain why and how they must be modied and combined in order to describe the behavior of very small objects. Our most complete theory of light is a quantum eld theory having both wavelike and particle-like propertiesthat is, incorporating wave-particle dualism. We will clarify the relationships between classical particles, classical electromagnetic waves and quantum eld theory by asking what each of them predicts for a single slit experiment. 1b. Classical Particles. A particle is characterized by having innitesimal size and a denite location in space at each instant of time. It may have mass and an electrical charge, it certainly has momentum and carries kinetic energy. As time increases, the particle moves along a denite trajectory, which is determined entirely by local conditions at the point where the particle is. For example, in Fig. 1 we have a particle with charge q moving near a charged plate. At each instant of time, the particle experiences a force F = q E determined by the electric eld at the location of the particle. The subsequent motion of the particle in the

next innitesimal interval of time is governed by: a = F /m = (q/m)E. If there are no forces acting on it, the particle travels in a straight line at constant speed. 1c. Classical Waves. The wave concept is abstracted from many dierent phenomena that are only geometrically similar, such as waves on the surface of water, sound in gas, liquid or solid, electromagnetic waves, displacement of a taut string or a stretched membrane (drumhead, apping sail) etc. In each case there is some eld quantity varying smoothly in 1, 2 or 3 dimensions. (For water waves, vibrating strings and membranes it is the displacement of particles from their equilibrium position; for sound in air it can be density or pressure variations; for electromagnetic waves there are six varying elds, the three components of E and the three components of B.) We will not need to consider the detailed electromagnetic wave picture for light that can be derived from Maxwells dierential equations of the electromagnetic eld, and which correctly predicts the lights velocity, its polarization (i.e., the fact that the E vector is perpendicular to the direction of propagation of the light wave) and the waves behavior at surface boundaries with conducting and dielectric materials. Rather, we will consider the most characteristic property of any wave, that it undergoes diraction at an obstacle. The momentum and energy carried by a wave are not concentrated at a point, but are spread out smoothly over a nite volume. This is in contrast to the very small sizes of atomic particles, and this contrast is what makes it hard to imagine an entity which is simultaneously a wave and a particle. In fact, we will see that light is composed neither of classical waves nor of classical particles, but is a new kind of entity that behaves in certain ways like a particle and in other ways like a wave. A probabilistic description ties these aspects together.

+ ++ + + ++ + + + ++ + +

Figure 1. Local inuences on the trajectory of a charged particle near a charged plate.

MISN-0-246

MISN-0-246

source

shutter

slit D

detector array

. . . 5 4 3 2 1

no. of events/bin
12345...

bins

Figure 2. Single slit diraction; experimental apparatus. 1d. The Nature of Light. Is light: A wave? Evidence: diraction and interference experiments. An electromagnetic wave? Evidence: Maxwells equations predict such behavior as its speed in vacuum, Snells law of refraction, polarization, generation of long waves by simple electromagnetic circuits, etc.

Figure 3. Particle Model, computer generated single-slit bar histogram; 41 detectors, 2305 particles. open the shutter for a certain time interval, then examine the distribution of light energy among the various detectors. We will assume that R L so the rays of light arrive at the slit as a parallel bundle. This allows the Fraunhfer condition to be fullled so o that the mathematical description is simpler. The detection of light takes place in a two dimensional array of individual, nite-sized detectors, each with a detection threshold E0 ; that is, a minimum amount of energy E0 must reach the detector in some time t in order that the detector registers the presence of light. Examples of such arrays include the eyes retina, with rods and cones; a photographic plate, with individual silver halide crystals as detectors (size typically 1 m); a stack of photomultiplier tubes; an array of thermocouples, sensitive to the heat energy from the light. Any of these could be used in the experiment. Now let the shutter open for a time interval T t, and look at the energy accumulated by each detector, as predicted by each of the models for light.

Evidence: Compton eect, photoelectric eect, photochemical reactions, etc. The seemingly contradictory wave and particle natures of light are reconciled by quantum eld theory. Its basic statement is: The intensity (at a given point in space and time) of an electromagnetic wave of frequency gives the probability, and only the probability, that energy in the amount h may be transferred between the wave and an external object at that point in space. The wave, particle and quantum eld theory ideas can be contrasted, compared and explained by using each to predict the results of a Fraunhfer single slit diraction experiment. o

A particle?

2. Both Aspects in One Experiment

2a. The Single Slit Apparatus. Lights seemingly contradictory wave and particle aspects can be made to appear simultaneously in the single-slit experiment of Fig. 2. The general idea of this experiment is to Figure 4. As in Fig. 3, but a much larger number of particles.

MISN-0-246

MISN-0-246

2b. Particle Model Predictions. We assume that all the light particles carry the same energy Ep , which is greater than the detector threshold E0 . To each detector we attach a counter, called a bin, which is advanced one unit by the detector when it detects a particle. In Fig. 3 the results of a computer calculation simulating the results of the experiment are shown. (Randomly spaced straight line paths of the particles from the source to the detectors were assumed). Shown on the graph is the number of particles detected by each of 41 detectors in a row across the array. A total of 2305 particles arrived, but only 16 detectors show any particles (an average of 144 each); the rest are in shadow. (The end detectors of the 16 were partly in and partly out of the shadow and, especially on the left, show fewer counts). The random spacing between trajectories accounts for the uctuations in the numbers of particles arriving at dierent detectors. Simple statistical theory predicts that the average uctuation, n, in number from one detector to another is given by n < n > where < n > is the average number of particles reaching any one detector in the time interval, t, that the shutter is open. As the total number of particles reaching the detectors becomes very large, this uctuation becomes comparatively unnoticeable, and so a strip of width D appears uniformly illuminated, with a sharp edged shadow, as shown in Fig. 4. This is the characteristic illumination pattern given by particle theory (ray, or geometrical optics theory), and in fact it represents rather well everyday (every sunny day at least) observations made under certain circumstances. (What are they?) 2c. Wave Model Predictions. Continuing the above experiment (but with real light) let us replace the detector array with one whose individual detectors are much smaller and closer together. We will notice that the boundary between shadow and light is not completely sharp, as predicted by the particle model, but has a width s L where is the wavelength of the light. This feature is explained by a wave model of light; but instead of pursuing it further, we will move on to a mathematically simpler case by shrinking D until it is much smaller than the width of the light-shadow boundary. Then the wave model predicts a Fraunhfer diraction pattern in which the light of intensity I smoothly o varies along the detector array according to the equation I(y) = I0 (sin2 )/ 2 , = Dy/L,

Figure 5. Fraunhfer ino tensity pattern for 41 detectors, 1405 energy units. Now we must look at the process of wave detection. For particles, our detectors counted. But wave theory assumes that the energy is distributed smoothly in space and time and one could give a continuous readout of the energy collected versus time. However, real detectors must have nite size and nite time resolution, and so we assume, again, that the detector must absorb energy until it collects an amount E0 . It then advances its bin counter by one unit and is reset to begin the accumulation of another amount E0 of energy. For an array of 41 detectors which has registered the arrival of 1405 units of energy, wave theory predicts that we get the Fraunhfer pattern of o Fig. 5. But if the array contains many more detectors, and the threshold E0 is very small, a smooth curve results (Fig. 6). 2d. Both Aspects Observed Simultaneously. Both the wave and the particle aspects of light will show up simultaneously in our single slit experiment if: 1. D L, so the diraction pattern is present.

2. the intensity is low enough so we can have one photon at a time. We also require that the detectors be sensitive enough that their Ep = h. That is, they can detect single quanta of light. threshold E0 With all bins empty, we open the shutter. The detectors begin to detect photons at random intervals of time. As the counts build up, however, we see neither the sharp-edged pattern appropriate for classical particles (Fig. 3) nor the smooth diraction pattern of a wave theory (Fig. 5) but

where y measures distance along the detector array from the center of the pattern.
9

Figure 6. As in Fig. 5, but more detectors, small threshold.

10

MISN-0-246

MISN-0-246

Figure 7. A non-particle, non-wave pattern, observable under appropriate circumstances. rather a noisy Fraunhfer pattern such as Fig. 7 shows after 1405 have o been detected.

} l/2
slit jaws A

}
screen
Figure 8. .

3. The Quantum Field Theory of Light

3a. Light Is Not Particles, Not Waves. Light cannot be categorized as either classical particles or classical waves. If it were classiable as a classical particle it would not produce a diraction pattern, but light diracts. If it were a classical wave, its energy and momentum would have a continuous distribution in space and time, but light often acts as if it consists of localized photons having discrete energy and momenta. Thus light cannot be said to be intrinsically one or the other. 3b. Quantum Field Theorys Description. The principle features of the quantum eld theory description of light may be stated thusly: 1. Electromagnetic eld components, E and B, still obey Maxwells equations. When crossed and sinusoidally uctuating, they still form an electromagnetic wave traveling at speed c. However, the E and B values no longer give exact forces on charged particles, hence no longer predict exact energy and momentum transfer rates to those particles. 2. An electromagnetic wave can only transfer energy and momentum to and from charged particles in increments xed by the waves frequency: E = h; p = h/c. 3. The actual times and places of these energy and momentum transfers cannot be predicted exactly, but the probabilities that they will occur in specied time and space intervals can be precisely calculated. These precise probabilities are linearly proportional to the waves electromagnetic eld intensities. The rst statement delineates wavelike properties of light; the second, particle-like properties. The third statement shows how we reconcile those
11

properties by including the observed random nature (unpredictableness) of the energy and momentum transfers. It allows us to (correctly) predict particle-like properties for small numbers of events and for short time intervals, and wavelike properties for large numbers of events where probabilistic distributions and actual distributions merge. Here are some more details on the ways each classical model fails for light: 1. Light is not a classical wave. For example, consider the photoelectric eect using a dust cathode. When light of low intensity is shown on ne metallic dust, photo-electrons start coming o immediately, whereas wave theory predicts that a long time must elapse before a dust particle can collect enough energy over its area from a wave to emit a photoelectron.1 2. Light is not a classical particle. That is, the photon cannot be thought of as having a straight-line trajectory of innitesimal width. Experiments designed to establish which part of the slit opening a given photon crossed (for example by having a second, narrower slit) will cause a further diraction or other disturbance of the photon and no denite path can be established. It is most correct to think of photons only at their instant of creation or destruction, and to consider light to be a probability wave in between these times, although in the geometrical limit (resolution of the path is less than L) an approximate trajectory can be assigned to the light wave, as though it were a particle. 3c. Classical EM Wave Theory. Figure 8 shows a slit with shiny metal jaws. A light wave is incident on the slit, its E vectors perpendicular to the page.
1 See

The Photoelectric Eect (MISN-0-213).

}
12

MISN-0-246 In region A the wave is propagating as in free space.

MISN-0-246

10

source

shutter

slits

detectors

In region B, where the wave is near the conducting metal, the time varying E eld causes a current to ow in the metal. This current produces a B(t) eld which produces (Faradays law) another E(t) eld. All these E(t) and B(t) elds add to yield nal E and B(t) elds that satisfy Maxwells equations for a wave dierent from the incident one. This is the diracted wave. In addition, the electrical resistance of the metal causes some of the light energy to be absorbed at the slit jaws. In region C, the currents produce the B(t) [and thus also the E(t)] elds of a reected wave; some light is also absorbed. The above description uses only the electromagnetic wave theory. All we need to add to the above is that energy is absorbed in terms of photons, each of energy h and transferring momentum h/c to the slit. (The reected photons transfer momentum 2h/c to the slit). These photons are absorbed at random places and times on the metal in proportion to the intensity of the light ( E 2 ) and to a transition probability that depends on the electrical properties of the metal (such as its resistance). Consider a little volume of light at region B in time interval t. There is a probability Pa that a photon will be absorbed by the slit jaws. Then there is a probability Pb = 1 Pa that no photon will be absorbed, in which case, the light in region B propagates down to the screen where new probabilities of the light being reected, transmitted and absorbed can be calculated, depending on E 2 there and on the electrical properties of the screen.

Figure 9. Apparatus for a 2-slit experiment. ton is created or destroyed in a small, localized region. Consider a gas discharge tube in the center of the lab, and have photomultiplier detectors on the opposite walls (Fig. 10). Assume that, on the average, one atom per second of the source emits a photon, considered a spherical electromagnetic wave pulse that expands in all directions with speed c. After a time T = r/c, either detector 1 or detector 2, or neither, will detect a photon containing all the energy of the emitted electromagnetic wave. In no case will detector 1 and detector 2 both detect parts of the same spherical wave pulse if the intensity is suciently low. 4c. Designing and Using an Astronomical Telescope. We can use the design and uses of an astronomical telescope to illustrate when it is appropriate to use the wave and when the particle approximations for light, and when it is necessary to use the more exact quantum eld theory:

4. Field Theory Applications

4a. Strange One- and Two-slit Predictions. A single photon must go through both slits of a double slit diraction apparatus (see Fig. 9). With low intensity light (say an average of 1 photon/sec) a distinctive double slit diraction pattern builds up as many photons are detected. If one slit is closed, the pattern becomes that of a single slit. It is dicult to explain this on a particle model; why would the use or non-use of one slit aect a particle through the other one? (The low intensity is specied to avoid possible cooperative eects between several photons and the slits, e.g. interactions between photons of one slit and those of the other.) 4b. Strange Spherical Wave Predictions. Although a classical electromagnetic wave may extend over a large volume of space, the pho-

PM

PM
Figure 10. Photomultipliers (PM) detecting a ash of light.

13

14

MISN-0-246

11

MISN-0-246

PS-1

1. Size of lenses and mirrors, their placement, image size, magnication, etc. Use light rays, geometrical optics, laws of reection and refraction (classical particle theory). 2. Resolving power of the telescope, diraction pattern of a point source (star) processed through a circular aperture (lens). Use wave theory. 3. Analysis of the light for information about distance and speed of the source, media through which the light passed (interstellar space); Doppler shift, etc. Use electromagnetic wave theory. 4. For very faint sources, the telescope may detect only a few hundred photons appearing at random points in a diraction pattern. Use quantum eld theory to make a statistical analysis of the pattern.

PROBLEM SUPPLEMENT
Note: these problems also occur on this modules Model Exam. 1. For each application given below, give reasons it can or cannot be satisfactorily described as each of these: (a) a classical particle; (b) a classical EM wave; and (c) a quantum eld. The Phenomena are: A. B. C. D. E. F. G. H. design of binoculars architectural analysis of light in and around a building study of bee navigation (sky polarization eects) analysis of light from very distant galaxies using a photomultiplier tube study of formation of vitamin D by suntanning using a light microscope to study bacteria 1m long installation of a magic eye door opener building a pin hole camera

Acknowledgments
We thank Professor James Linneman for helpful suggestions. Preparation of this module was supported in part by the National Science Foundation, Division of Science Education Development and Research, through Grant #SED 74-20088 to Michigan State University.

Glossary
quantum eld theory: a theory that reconciles the wave and particle duality of light by stating that the intensity of an electromagnetic wave at a given point in space is related to the probability of an energy and momentum transfer by a photon. wave-particle duality: a model that encorporates both wave-like and particle-like properties.

2. An atom emitting light in the visible typically takes about 108 seconds to do it. How long is the resulting wave packet in space?

Brief Answers: 1. A. a, geometrical optics B. C. D. E. F. G. H. a, geometrical optics b, polarized EM wave c, photon statistics c, quantum of light energy converted to bond energy b, size of diraction eects c, photon, photo-electric eect a and b

2. About 3 meters.

15

16

MISN-0-246

ME-1

MODEL EXAM
1. See Output Skills K1-K2 on this modules ID Sheet. One or more of these skills, or none, may be on the actual exam. 2. For each application given below, give reasons it can or cannot be satisfactorily described as each of these: (a) a classical particle; (b) a classical EM wave; and (c) a quantum eld. The Phenomena are: A. design of binoculars B. architectural analysis of light in and around a building C. study of bee navigation (sky polarization eects) D. analysis of light from very distant galaxies using a photomultiplier tube E. study of formation of vitamin D by suntanning F. using a light microscope to study bacteria 1m long G. installation of a magic eye door opener H. building a pin hole camera 3. An atom emitting light in the visible typically takes about 108 seconds to do it. How long is the resulting wave packet in space?

Brief Answers: 1. See this modules text. 2. See problem 1, this modules Problem Supplement. 3. See problem 2 this modules Problem Supplement.

17

18

19

20