2.91 Optics: Lecture 11 Notes

Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang Outline: A. Superposition of waves, Interference B. Interferometry - Amplitude-splitting (e.g. Michelson interferometry) - Wavefront-splitting (e.g. Young’s Double Slits) C. More Interferometry: Fabry Perot, etc A. Superposition of Waves, Interference The nature of linear wave equation guarantees that waves can be superimposed: we may combine an array of waves by algebra, as far as each of them are proper solution of the wave equation. To facilitate this process, we make use of the following complex number to represent the wave field. - Waves in complex numbers For example, the electric field of a monochromatic light field can be expressed as: 𝐸(𝑧, 𝑡) = 𝐴𝑐𝑜𝑠(𝑘𝑧 − 𝜔𝑡 + 𝜑) (1) Since exp(𝑖𝑥) = cos(𝑥) + 𝑖 sin(𝑥) (2) 𝐸(𝑧, 𝑡) = 𝑅𝑒{𝐴𝑒𝑥𝑝[𝑖(𝑘𝑧 − 𝜔𝑡 + 𝜑)]} (3) 𝐸(𝑧, 𝑡) = 2 {𝐴𝑒𝑥𝑝[𝑖(𝑘𝑧 − 𝜔𝑡 + 𝜑)]} + 𝑐. 𝑐. (𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑐𝑜𝑛𝑗𝑢𝑔𝑎𝑡𝑒) (4) Or 1 - Complex numbers simplify optics! Interference with two beams e.g. 2 plane waves propagating in +z direction 𝐸1𝑥 = 𝐸1𝑥 (0)𝑒𝑥𝑝[𝑖(𝑘1 𝑧 − 𝜔1 𝑡 + 𝜑1 )] 𝐸2𝑥 = 𝐸2𝑥 (0)𝑒𝑥𝑝[𝑖(𝑘2 𝑧 − 𝜔2 𝑡 + 𝜑2 )] (5) (6) We can define phase of each waves: 𝛿1 (𝑧, 𝑡) = 𝑘1 𝑧 − 𝜔1 𝑡 + 𝜑1 𝛿2 (𝑧, 𝑡) = 𝑘2 𝑧 − 𝜔2 𝑡 + 𝜑2 (7) (8) For a point P located at z=z0 the combined field is: 𝐸𝑥 = 𝐸1𝑥 + 𝐸2𝑥 = 𝐸1𝑥 (0) exp[𝑖𝛿1 (𝑧0 , 𝑡)] + 𝐸2𝑥 (0) exp[𝑖𝛿2 (𝑧0 , 𝑡)] (9) 1 Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang B. Interferometry We often use optical interferometers to facilitate the study of interference. A few common setups are discussed here. Based on the operation principle to split the beams, we may find the so-called amplitude-splitting or wavefront-splitting devices. - Input beam E(t) Mirror Michelson Interferometry The Michelson Interferometer is named after Albert Michelson, who used it with Edward Morley in 1887, in an attempt to measure the existence of the "ether". Output beamE(t–t) Translation stage In Michelson- Morley’s famous experiment, the delay time ∆𝜏 is achieved simply by moving a mirror along the optical axis. Moving a mirror backward by a distance L yields a delay of: ∆𝜏 = 2𝐿 (10) 𝑐 (e.g. 300 µm of mirror displacement yields a delay of 2 × 10−12 s=2ps). Input beam L2 Mirror I0 Beamsplitter Output beam I0 L1 Delay A Michelson Interferometer as shown in left schematic, split a beam of incident light into two arms using a thin glass window. Both beams travel to mirrors that are precisely aligned to reflect them. Before recombining them at the beam splitter, the two beams traveled with different optical path length L1 and L2. 𝐼 = 2𝐼0 + 2𝐼0 〈cos (2𝜔 L1 −L2 Mirror 𝑐 )〉 (11) The variation of intensity as a function of the path length L1 gives us a measure of wavelength of light! Recent effort is to apply such technology in measurement of gravity waves. Observation: - Michelson Interferometer measures (auto)-correlation in time. To see this effect, we suppose the input light beam is not monochromatic. Thus 𝑐 𝑐 ∗ ∗ 𝐼 = 2 𝜀〈𝐸𝑥 ∙ 𝐸𝑥∗ 〉 = 2 𝜀〈(𝐸1𝑥 + 𝐸1𝑥 (𝑡 − 𝜏)) ∙ (𝐸1𝑥 + 𝐸1𝑥 (𝑡 − 𝜏))〉 (12) 2 Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang ∗ 𝐼 = 𝐼1 (𝑡) + 𝐼1 (𝑡 − 𝜏) + 2〈𝐸1𝑥 ∙ 𝐸1𝑥 (𝑡 − 𝜏)〉 (13) Such auto-correlation function tells the similarity of the field over given period of time. The Fourier transform of the auto-correlated signal yields the Power Spectrum: 𝐹𝑇 ∞ ∗ ∗ ∫0 𝐸1𝑥 ∙ 𝐸1𝑥 (𝑡 − 𝜏)𝑑𝑡 ⇒ 𝐸1𝑥 (𝜔) ∙ 𝐸1𝑥 (𝜔) = |𝐸1𝑥 (𝜔)|2 (14) This is how Fourier Transform Spectroscopy (often abbreviated as FTIR) are constructed nowadays. (See Pedrotti 21-2 for more discussion) Interferogram This interferogram is very narrow, so the spectrum is very broad. 4000 3200 2400 1600 800 © Source unknown. All rights reserved. © Source unknown. All rights reserved. - In the above analysis we assumed the input beams are ideal plane waves so we only focused on intensity variation of a single spot as the arm length changes. In reality, a set of nested rings are often observed on the receiving screen as the pattern on right. This is due to the variation of phase as a function of momentum difference (𝑘1 − 𝑘2 )𝑧. - Young’s Double Slit Interferometry X X’ observation point x’ x=x0 Z x=−x0 incoming plane wave (on-axis) opaque screen To analyze Young’s experiment, we assume the screen X with two narrow slit is illuminated with a monochromatic plane wave. After the slits, two cylindrical waves are excited. This creates fringes at the observation plane X’ , after travelling a distance 𝑧 = 𝑙. z =l the phase difference of the two waves becomes: For a position x’ on the screen, 3 Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang 𝛿1 − 𝛿2 = 𝑘(𝑟1 − 𝑟2 ) + (𝜑1 − 𝜑2 ) (15) where 𝑟1 = √(𝑥 ′ − 𝑥0 )2 + 𝑙 2 , 𝑟2 = √(𝑥 ′ + 𝑥0 )2 + 𝑙 2 . When the slits are arranged symmetrically and the incoming plane wave is at normal incidence, then 𝜑1 − 𝜑2 = 0 In a typical experiment the screen is placed far away from the slits (𝑙 ≫ 𝑥0 ) so we can further take approximation: 𝑟1 = 𝑙√1 + 𝑟2 = 𝑙 √1 + (𝑥 ′ −𝑥0 )2 𝑙2 (𝑥 ′ +𝑥0 )2 𝑙2 ≈ 𝑙 (1 + ≈ 𝑙 (1 + (𝑟1 − 𝑟2 ) = − (𝑥 ′ −𝑥0 ) 2 2𝑙 2 (𝑥 ′ +𝑥0 ) 2𝑙 2 2𝑥′𝑥0 2 ) (16) ) (17) (18) 𝑙 In order for the constructive interference to occur, cos(𝛿1 − 𝛿2 ) = 1, then 𝑘(𝑟1 − 𝑟2 ) = −𝑘 2𝑥0 𝑥 ′ 𝑙 = 2𝑚𝜋 (19) 𝑚 = 0, ±1, ±2, …. are called order number. This gives rise to the famous condition: 2𝑥0 𝑥 ′ = 𝑚𝑙𝜆 (20) Therefore Young’s experiments directly measured the wavelength of light (in 18 century!) Note: Generally fringes will form by two or more beams crossing at an angle. To quantify that we can modify our phase term for the crossing beams: 𝛿1 (𝑥, 𝑧, 𝑡) = 𝑘1𝑥 𝑥 + 𝑘1𝑧 𝑧 − 𝜔1 𝑡 + 𝜑1 or in terms of incident angle θ: 𝛿1 (𝑥, 𝑧, 𝑡) = 𝑘1 𝑥𝑠𝑖𝑛𝜃1 + 𝑘1 𝑧𝑐𝑜𝑠𝜃1 − 𝜔1 𝑡 + 𝜑1 𝛿1 − 𝛿2 = (𝑘1 𝑠𝑖𝑛𝜃1 − 𝑘2 𝑠𝑖𝑛𝜃2 )𝑥 + (𝑘1 𝑐𝑜𝑠𝜃1 − 𝑘2 𝑐𝑜𝑠𝜃2 )𝑧 + (𝜑1 − 𝜑2 ) (21) (22) (23) Fringes can vary both on x and z directions! Note: Young’s double slit experiment measures the correlation in space. Assuming the input light beam is not a plane wave (i.e. inhomogeneous), at x’=0 we measure the interference: 𝑐 𝑐 ∗ ∗ 𝐼 = 2 𝜀〈𝐸𝑥 ∙ 𝐸𝑥∗ 〉 = 2 𝜀〈(𝐸1𝑥 (𝑥 − 𝑥0 ) + 𝐸1𝑥 (𝑥+𝑥0 )) ∙ (𝐸1𝑥 (𝑥 − 𝑥0 ) + 𝐸1𝑥 (𝑥 + 𝑥0 ))〉 (24) - 4 Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang ∗ (𝑥 + 𝑥0 )〉 𝐼 = 𝐼1 (𝑥 − 𝑥0 ) + 𝐼1 (𝑥 + 𝑥0 ) + 2〈𝐸1𝑥 (𝑥 − 𝑥0 ) ∙ 𝐸1𝑥 (25) Such correlation function tells the similarity of the field over a given spatial period. This effect is often used to measure the coherence of a remote star under the telescope, although the radiation is thought to be randomly distributed. A daily life example is the spatial coherence of ripples in the pool (“Spatial coherence from Ducks” by Emil Wolf et al, Physics today 2010). - Comparison between Michelson and Young’s double slits: Both can be regarded as interference of two spherical waves, but observed in different directions. Image of wavefront splitting inferometry removed due to copyright restrictions. - Other wavefront splitting interferometry similar to Young’s Double slits: o Lloyd’s Mirror o Fresnel’s biprism o Fresnel’s mirror o Billet’s split lens C. More examples of Interferometry - Thin Film interference 5 Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang - Fabry-Perot Interferometry t’(rr’)2tE0e2id 2tE r’(rr’) 0 Consider two parallel reflective tr’(rr’)tE0e2id surfaces separated by distance d, (rr’)2tE0 t’rr’tE0eid the first one has a amplitude r’rr’tE 0 tr’tE0eid transmission and reflection rr’tE0 coefficient t and r, and the r’tE0 t’tE0 second has amplitude rE0 transmission and reflection tE0 E0 coefficient t’ and r’. Multiple r, t r’, t’ reflections between the two surfaces results in two series of reflected and transmitted terms. Due to the round trip travel path, there is a phase difference between successive transmitted terms: (26) 𝛿0 = (2𝑘𝑧 𝑑) The transmitted series is 𝐸𝑡 = exp(𝑖𝛿0 /2)(𝑡 ′ 𝑡𝐸0 + 𝑡 ′ (𝑟𝑟 ′ )𝑡𝐸0 exp(𝑖𝛿0 ) + 𝑡 ′ (𝑟𝑟 ′ )2 𝑡𝐸0 exp(2𝑖𝛿0 ) + ⋯ ) ′ 𝐸𝑡 = 𝑡 𝑡𝐸0 exp(𝑖𝛿0 /2)[1 + 𝐸𝑡 = 𝑡 ′ (𝑟𝑟 ′ ) exp(𝑖𝛿0 ) + (𝑟𝑟 ′ )2 exp(2𝑖𝛿0 ) + ⋯ ] ′ 𝑛 𝑡𝐸0 exp(𝑖𝛿0 /2) ∑∞ 𝑛=0(𝑟𝑟 ) exp(𝑖𝑛𝛿0 ) 𝐸𝑡 = 𝑡 ′ 𝑡exp(𝑖𝛿0 /2)𝐸0 (27) (28) (29) (30) 1−𝑟𝑟′ exp(𝑖𝛿0 ) The transmitted irradiance is given by: 𝑇′𝑇𝐼0 𝐼𝑡 = |1−𝑟𝑟′ (31) exp(𝑖𝛿0 )|2 The denominator of the last result can be expressed as: |1 − 𝑟𝑟′ exp(𝑖𝛿0 )|2 = (1 − 𝑟𝑟′ exp(𝑖𝛿0 ))(1 − 𝑟 ∗ 𝑟 ′∗ exp(−𝑖𝛿0 )) = 1 − (𝑟 𝑟 ′ exp(𝑖𝛿0 ) + 𝑟 ∗ 𝑟 ′∗ exp(−𝑖𝛿0 )) + 𝑅𝑅′ = 1 − 2√𝑅𝑅 ′ cos(𝛿) + 𝑅𝑅′ 2 = (1 − √𝑅𝑅 ′ ) + 2√𝑅𝑅 ′ (1 − cos(𝛿)) 2 𝛿 = (1 − √𝑅𝑅 ′ ) + 4√𝑅𝑅 ′ sin2 (2) Where 𝛿 = (2𝑘𝑧 𝑑) + 𝜙𝑟 + 𝜙𝑟′ (32) (33) We define the coefficient of finesse F: 6 Lecture Notes on Wave Optics (03/12/14) 2.71/2.710 Introduction to Optics –Nick Fang 4√𝑅𝑅 ′ ℱ= (34) 2 (1−√𝑅𝑅 ′ ) To express the general form of It: 𝐼𝑡 = 𝑇′𝑇𝐼0 2 (1−√𝑅𝑅 ′ ) [ 1 𝛿 2 1+ℱsin2 ( ) ] (35) Applications: Fabry-Perot cavities are often designed to distinguish closely spaced spectral lines of a gas medium. Higher values of Finesse F give a sharper transmission pass band and greater spectral resolution. To find the half-width of the pass band, we solve: 1 giving 𝛿 1+ℱsin2 ( ) 2 𝛿1/2 𝑠𝑖𝑛 ( 2 = )= 1 2 1 √ℱ (36) (37) Figure 8.9 from Pedrotti: Transmittance of Fabry-Perot Cavity. © Pearson Prentice Hall. All rights reserved. 7