## Trova il tuo prossimo libro preferito

# Computer Design of Diffractive Optics

Di V A Soifer

## Descrizione

The theoretical aspects underpinning diffractive optics are initially explored, including the main equations in diffraction theory and diffractive optical transformations. Application of electromagnetic field theory for calculating diffractive gratings and related methods in micro-optics are discussed, as is analysis of transverse modes of laser radiation and the formation of self-replicating multimode laser beams. Key applications of DOEs reviewed include geometrical optics approximation, scalar approximation and optical manipulation of micro objects, with additional consideration of multi-order DOEs and synthesis of DOEs on polycrystalline diamond films.

With its distinguished editor and respected team of expert contributors, Computer design of diffractive optics is a comprehensive reference tool for professionals and academics working in the field of optical engineering and photonics.

- Explores the theoretical aspects underpinning diffractive optics
- Discusses key applications of diffractive optical elements
- A comprehensive reference for professionals and academics in optical engineering and photonics

## Informazioni sull'autore

## Autori correlati

## Correlato a Computer Design of Diffractive Optics

Fundamentals of Femtosecond Optics di S A Kozlov Valutazione: 0 su 5 stelle

## Categorie correlate

## Anteprima del libro

### Computer Design of Diffractive Optics - V A Soifer

**Woodhead Publishing Series in Electronic and Optical Materials **

**Computer design of diffractive optics **

V.A. Soifer

**Table of Contents **

**Cover image **

**Title page **

**Copyright **

**Preface **

**Chapter 1: Main equations of diffraction theory **

**1.1 Maxwell equations **

**1.2 Differential equations in optics **

**1.3 Integral optics theorems **

**1.4 Integral transforms in optics **

**1.5 Methods for solving the direct diffraction problem **

**Conclusion **

**Chapter 2: Diffractive optical transformations **

**2.1 Transformations in optical systems **

**2.2 Diffraction gratings **

**2.3 Flat lenses and prisms **

**2.4 Inverse problem of diffractive optics **

**2.5 The method of coding the phase function of DOE **

**2.6 Discretisation and quantisation of the DOE phase **

**2.7 Computer design and formation of the diffractive microrelief **

**Chapter 3: Calculation of diffractive optical elements in geometrical optics approximation **

**3.1 Calculation of DOE for focusing into a curve in geometrical optics approximation **

**3.2 Curvilinear coordinates in the problem of focusing on a curve **

**3.3 Calculation and investigation of geometrical optics focusators **

**3.4 Focusator into a two-dimensional region. The method of matched rectangles **

**3.5 Correction of wave fronts **

**Conclusion **

**Chapter 4: Calculation of the DOE in the scalar approximation of the diffraction theory **

**4.1 Iterative methods of calculating the DOE **

**4.2 Calculation of the DOEs producing the radial-symmetric intensity distribution **

**4.3 Calculation of one-dimensional diffractive gratings **

**4.4 The equalisation of the intensity of the Gaussian beam **

**4.5 DOE forming contour images **

**4.6 Calculation of quantised DOEs **

**Conclusions **

**Chapter 5: Multi-order diffractive optical elements **

**5.1 Multi-order focusators **

**5.2 Diffractive multi-focus lenses **

**5.3 Two-order DOEs **

**5.4 Spectral DOEs **

**Conclusions **

**Chapter 6: Application of the theory of the electromagnetic field for calculating diffractive gratings **

**6.1 Diffraction on ideally conducting gratings with a stepped profile **

**6.2 Diffraction on the ideally reflecting gratings with a continuous profile (Rayleigh approximation) **

**6.3 Diffraction on dielectric gratings **

**6.4 Gradient methods of calculating the profile of the diffractive gratings **

**6.5 Diffraction on two-dimensional dielectric gratings **

**Conclusions **

**Chapter 7: Methods of the theory of the electromagnetic field in micro-optics **

**7.1 Analysis of the DOE by the method of finite-difference time-domain solution of Maxwell equations **

**7.1.3 Diffraction of the TE mode on the two-dimensional gratings with ideal conductivity **

**7.2 The finite element method in micro-optics **

**Chapter 8: Analysis of transverse modes of laser radiation **

**8.1 Propagation of electromagnetic radiation in optical waveguides **

**8.2 Modans – diffractive optical elements (DOE) matched to laser radiation modes **

**8.3 Calculation of the DOE matched with the characteristics of the gradient medium **

**8.4 DOEs for analysis of the transverse modes of light fields **

**8.5 Selection of modes in free space **

**8.6 Transmission of information with mode-division multiplexing **

**8.7 Fibre optic sensors based on mode selection **

**Chapter 9: Formation of self-reproducing multimode laser beams **

**9.1 Multimode light fields with different properties of self-reproduction **

**9.2 Composition method for the synthesis of DOE forming a multimode beam **

**9.3 Formation of self-reproducing multi-mode laser beams **

**9.4 Formation of several self-reproducing beams in different diffraction orders **

**Conclusion **

**Chapter 10: Optical manipulation of micro–objects by DOE **

**10.1 The strength of interaction of the light field with micro–objects **

**10.2 Light beams to capture micro–objects **

**10.3 The scope of optical manipulation **

**10.4 Motion control of micro–objects using light fields formed by DOE **

**Conclusion **

**Chapter 11: Synthesis of DOE on polycrystalline diamond films **

**11.1 Formation technology of the microrelief on the surface of diamond films **

**11.2 Synthesis and study of thin lenses on diamond films **

**11.3 DOEs focusing CO2-laser radiation in two-dimensional field **

**11.4 Analysis of antireflective subwavelength structures formed on the diamond film **

**11.5 Simulation of a cylindrical diamond DOE with subwavelength technological errors in the microrelief **

**11.6 The influence of local technological errors on efficiency of the DOE **

**11.7 Stochastic optimization of the diamond focuser microrelief taking into account the systematic errors of manufacture **

**11.8 Experimental study of the focuser into a circle **

**Index **

**Copyright **

Published by Cambridge International Science Publishing Limited in association with Woodhead Publishing Limited

Cambridge International Science Publishing Limited, 7 Meadow Walk, Great Abington, Cambridge CB21 6AZ, UK

**www.cisp-publishing.com **

Woodhead Publishing Limited, 80 High Street, Sawston, Cambridge CB22 3HJ, UK

**www.woodheadpublishing.com; www.woodheadpublishingonline.com **

Woodhead Publishing, 1518 Walnut Street, Suite 1100, Philadelphia, PA 19102-3406, USA

Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road, Daryaganj, New Delhi - 110002, India

First published 2013, Cambridge International Science Publishing Limited and Woodhead Publishing Limited. This work is a revised and translated version of an original Russian-language version published in 2009 by Fizmatlit, Russia.

© Cambridge International Science Publishing Limited, 2013. The publishers have made every effort to ensure that permission for copyright material has been obtained by authors wishing to use such material. The authors and the publishers will be glad to hear from any copyright holder it has not been possible to contact.

The authors have asserted their moral rights.

This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book.

Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Cambridge International Science Publishing Limited and Woodhead Publishing Limited.

The consent of Cambridge International Science Publishing Limited and Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Cambridge International Science Publishing Limited and Woodhead Publishing Limited for such copying.

Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library.

Library of Congress Control Number: 2012948938

ISBN 978-1-84569-635-1 (print)

ISBN 978-0-85709-374-5 (online)

ISSN 2050-1501 Woodhead Publishing Series in Electronic and Optical Materials (print)

ISSN 2050-151X Woodhead Publishing Series in Electronic and Optical Materials (online)

The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elemental chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.

Typeset by Butterfly Info Services, India

Printed and bound in the UK by the MPG Books Group

**Preface **

In 2000, the publishing house Fizmatlit, Moscow, published the book Methods of Computer Optics, translated into English in 2002 and published by John Wiley & Sons, New York. A second, revised edition was published in 2003, which shows the interest from readers.

During this period, the main authors of the first edition expressed a critical approach to the presentation of some of the earlier chapters written by them and decided to improve them. Secondly, a lot of new research results in the field of electromagnetic methods of analysis and synthesis of diffractive optical elements with greater functionality and applications in laser systems appeared.

These reasons have prompted the authors to write a new book, based on the material of the previous editions, with expansion of the sections devoted to the solution of direct and inverse problems of diffraction theory. At the same time, given the limited scope of the content of the new book, many application issues, in particular issues of technology for diffractive optical elements, are described only concisely.

The book consists of 11 chapters, united by the idea of computer synthesis of diffractive optical elements with a highly functional conversion of laser radiation and the properties of the wave fields resulting from such changes.

The book was written by the Institute of Image Processing Systems, Russian Academy of Sciences: **Chapter 1 - D.L. Golovashkin, V.V. Kotlyar, Chapter 2 –V.A V.A. Soifer, Chapter 3 - L.L. Doskolovich, N.L. Kazan, V.A. Soifer, Chapter 4 - V.V. Kotlyar, L. Doskolovich, V.A. Soifer, Chapter 5 - L.L. Doskolovich, V.A. Soifer, Chapter 6 - L.L. Doskolovich, Chapter 7 - D.L. Golovashkin, V.V. Kotlyar, Chapter 8 - V.S. Pavelieva, S.N. Khonina, V.V. Kotlyar, V.A. Soifer, Chapter 9 - S.N. Khonina, V.V. Kotlyar, V.A. Soifer, Chapter 10 - R.V. Skidanova, S.N. Khonina, V.V. Kotlyar, V.A. Soifer, Chapter 11 - V.S. Pavelieva, D.L. Golovashkin, V.A. Soifer. **

The book is based on research carried out in the framework of the Rus- sian-American program ‘Basic Research and Higher Education’ (grant CRDF REC-SA-014-02), Basic Research (grants 04-01-96517, 04-07-90149, 04 -0208094, 05-01-08020, 05-08-50298, 05-08-04033), the President ofthe Russian Federation grant (MD-209.2003.01, 210.2003.01 MD, MD-4003.2005.09, MK-2568.2005.09, SS-1007.2003.01).

The authors would like to thank:

–D.V. Nesterenko, the Laboratory of Laser Measurements, The Image Processing Systems Institute of the Russian Academy of Sciences for, conducting computer simulations;

–Members of the Institute of General Physics, Russian Academy of Sciences, V.I. Konov and V.V. Kononenko for materials for **Chapter 11 of this book; **

–Foreign partners – FIAT Research Centre (the group of Dr. P. Perlo), the Institute of Applied Optics, Friedrich-Schiller University, Jena, Germany (teams of Prof. R. Kovarshik and Dr. M. Dyuppare), University of Joensuu, Finland (group of Prof. J. Turunen) for assistance in preparing the samples and the joint DOE experimental studies;

The book is based on original research papers published in the last 25 years by co-authors, and other included Academician A.M. Prokhorov, Prof. I.N. Sisyakin, M.A. Golub, A. Volkov, S.V. Karpeev and V.A. Danilov.

**1 **

**Main equations of diffraction theory **

**1.1 Maxwell equations **

**1.1.1 Mathematical concepts and notations **

The Hamiltonian operator in the Cartesian coordinate system is determined as

where **e***x*, **e ***y*, **e***z *are the unit vectors of the Cartesian coordinate system. The operators grad, div, rot and ∆ are defined as follows:

Where *f*, **F** = (*Fx*, *Fy*, *Fz*) are the scalar and vector functions, (.,.) and [.,.] are the operations of the scalar and vector products, det *A *is the determinant of the matrix *A*.

The most important integral relationships of vector analysis

The Ostrogradskii–Gauss theorem:

where **n **is the unit vector of the external normal, *V *is the domain of the space, restricted by the surface *S*.

The Stokes theorem

Here *L *is the contour, restricting the surface S.

**1.1.2 Maxwell equations in the differential form **

The electromagnetic theory of light is based on the system of Maxwell equations [**1] (in the Gauss unit system): **

**(1.1) **

**(1.2) **

**(1.3) **

**(1.4) **

The notations used here and in the rest of the book are presented in **Table 1.1. **

**Table 1.1 **

**Electromagnetic quantities **

The functions **E** = **E **(**r**, *t*), **H** = **H **(**r**, *t*), **D** = **D **(**r**, *t*), **B** = **B **(**r**, *t*) describe the electromagnetic field in the medium characterised by the parameters ε = ε (**E**, **r**, *t*), μ = μ (**H**, **r**, *t*), ρ = (**r**, *t*), **j** = **j **(**E**, **r**, *t*) (*r *are the spatial coordinates, *t *is time) and secondary current **j**sec which will be described separately.

Assuming that the processes in the medium are local and inertialess (the state at every point is independent of the adjacent points and at every moment of time is independent of ‘prior history’), the characteristics of the field and the medium are be linked by material equations [**1]: **

**(1.5) **

**(1.6) **

**(1.7) **

and by the law of charge conservation

**(1.8) **

Further, it is assumed that the parameters of the medium are independent of the vectors of the field and do not change with time: ε = ε (**r**), μ = μ (**r**) (linear medium), and are scalar (isotropic medium).

If the strength of the electrical and magnetic fields can be described in the form: **E** = Re(**E **exp (–*i*ω*t*)), **H** = Re (**H **exp(–*i*ω*t*)), where **E** = **E **(**r**), **H** = **H**(**r**) are the complex-valued functions [**1], ω is cyclic frequency, i is the apparent unity, we are concerned with a monochromatic field for which the equations (1.1), (1.2) take the following form: **

**(1.9) **

**(1.10) **

is the wavenumber.

**1.1.3 Maxwell integral equations **

Integrating the equations **(1.1), (1.2) with respect to the surface S restricted by contour L and accepting the Stokes theorem, we obtain the following equations: **

**(1.11) **

**(1.12) **

Equations **(1.3), (1.4) are integrated with respect to volume V restricted by the surface S. Subsequently, using the Ostrogradskii–Gauss theorem: **

**(1.13) **

**(1.14) **

The system **(1.11)–(1.14) is referred to as the Maxwell integral equations in the integral form. **

**1.1.4 Boundary conditions **

Using the Maxwell integral equations for the infinitely small contours and volumes at the interface of two media, we obtain the following boundary conditions [1) for the characteristics of the electromagnetic field:

**(1.15) **

**(1.16) **

**(1.17) **

**(1.18) **

is the density of surface current (plane, separating the media 1 and 2, is normal to the vector **e***y*).

**1.1.5 Poynting theorem **

The equation **(1.1) is multiplied by E and equation (1.2) by H and, consequently, we obtain: **

Deducting the second equation from the first one, we obtain the Poynting theorem [**1] according to which: **

**(1.19) **

In the integral form

**(1.20) **

we have the balance equation of the energy of the electromagnetic field in the volume *V*. The energy in the volume *V *is the Umov–Poynting vector, indicating the direction of movement of energy and is equal to the density of the energy flux.

The monochromatic field can be described by the Umov–Poynting complex vector

where the symbol * indicates the complex conjugation, and the mean value of the Umov–Poynting vector is equal to the real part of the complex vector.

**1.2 Differential equations in optics **

**1.2.1 Wave equations **

The current and charges, usually not found in optics problems, are excluded from the Maxwell equations. Consequently, equations **(1.1) and (1.2) take the following form: **

**(1.21) **

**(1.22) **

Dividing both parts of equation **(1.22) by μ and using the rot operator: **

**(1.23) **

Equation **(1.21) is differentiated with respect to time to exclude the second term from equation (1.23): **

Consequently, taking into account that

rot α**u **= α rot **u** + [grad α, **u**] and rot rot **u** = grad div **u **– V²**u **

we obtain:

**(1.24) **

For the equation div (ε**E**) = 0 we used the identity div α**u **= α div **u** + (**u**, grad α) and obtain ε div **E** + (**E**, grad ε) = 0. Expressing div **E **from the last equality, we substitute the result into **(1.24), writing the wave equation [1] for the strength of the electrical field in an inhomogeneous dielectric medium: **

**(1.25) **

The same procedure is used for deriving the wave equation for the strength of the magnetic field **H**:

**(1.26) **

For a homogeneous medium, electrical and magnetic μ permittivitties are constant and the wave equations take the form:

**(1.27) **

**(1.28) **

**1.2.2 Helmholtz equations **

The wave equations, written for the complex amplitudes (monochromatic waves), are referred to as the Helmholtz equations. For an inhomogeneous medium, these equations have the form:

**(1.29) **

**(1.30) **

and for a homogeneous medium:

**(1.31) **

**(1.32) **

The equations **(1.31) and (1.32) can be solved independently for every projection of the strength of the electrical and magnetic fields E and H, and these projections can be denoted by a single scalar function U: **

**(1.33) **

**1.2.3 The Fock-Leontovich equation **

The function *U *is represented in the form *U* = *U *exp (*ik*0 z) and the function is substituted into equation **,we obtain the Fock-Leontovich parabolic wave equation **

**(1.34) **

The parabolic equation **(1.34) in the scalar diffraction theory is used for describing light fields in a paraxial region which propagate mainly along some direction in the space in a small body angle. **

**1.2.4 Eikonal and transfer equations **

The function *U *is written in the form *U* = *U*0exp (*ik*0ψ), where ψ = ψ(*x*, *y*, *z*) is the eikonal, *U*0 is the amplitude (real function). Substituting this function into equation **(1.33) we obtain **

Equating the apparent part zero, we obtain the transfer equation:

**(1.35) **

The remaining terms form the following equation:

from which, setting λ→0 (approximation of geometrical optics), we obtain the eikonal equation:

**(1.36) **

is the refractive index of the medium.

**1.3 Integral optics theorems **

The characteristics of the electromagnetic field can be determined not only using the differential equations derived by Maxwell, Helmholtz and others, but also using integral equations and transforms equivalent to the former. The Maxwell equations for the monochromatic lightwave in the homogeneous region of space are equivalent the Stratton–Chu vector integral equations, the solution of the Helmholtz differential equation can be studied conveniently using the Kirchhoff–Helmholtz integral equation (Green’s third formula), and the Fock-Leontovich equation which holds in the paraxial regions, is equivalent to the Fresnel integral transformation.

**1.3.1 Green equations **

Two continuous functions *u *(*x, y, z*) and *v *(*x*, *y*, *z*) together with their derivatives in region *V *restricted by the piecewise-smooth surface *S*, are described by the second Green equation

**(1.37) **

where **n **is the Laplace operator or the Laplacian.

Using the Green equation **(1.37), the solution of the Helmholtz equation in the internal points of the homogeneous region V can be expressed by the values of the solution and its derivatives at the boundary S using the third Green equation (Helmholtz-Kirchhoff integral [2]): **

**(1.38) **

where *R *is the distance between the points **x**∈ *V*and **x'**∈ *S*, u(*x*,*y*,*z*) is the solution of the Helmholtz equation in the homogeneous space:

(Δ + *k *²) *u *(*x*, *y*, *z*) = 0

*k* = 2π/λ is the wavenumber of the wave with length λ.

The function

**(1.39) **

describes the spherical wave which is the Green function of the homogeneous space and satisfies the inhomogeneous Helmholtz equation with a point source:

**(1.40) **

In the regions of the space with a constant refractive index and without sources, the integral representation **(1.38) holds for any Cartesian component of the strength of the electrical field: **

**(1.41) **

*Diffraction of a scalar wave on a dielectric object *

*Diffraction of a scalar wave on a dielectric object*

As an example, we consider the scalar problem of diffraction of an electromagnetic wave on a homogeneous dielectric object [**3]. Let the functions E1(x) and E2(x) satisfy two Helmholtz equations inside the region V (inside the object) and on the outside: **

**(1.42) **

they satisfy the boundary conditions

**(1.43) **

and the Zommerfeld radiation conditions at infinity:

**(1.44) **

where *o*(*x*) is the function whose order of smallness is greater than *x *at *x *→ 0.

In the equations **(1.42) the function g describes the density of light sources outside the region V which the object occupies. There are no sources inside the object. Using the Green theorem (1.37) and (1.38), the solution of the system of equations (1.42), taking into account the conditions (1.43) and (1.44), can be reduced to the solution of the integral Fredholm equation of the second kind: **

**(1.45) **

where points **x **and **x**′ belong to the region of the object *V*, and point *y *belongs in the region *V'*, which is external in relation to region *V. *The second term in the equation **(1.45) describes the complex amplitude of the light field, incident on the object, which can be regarded as the known function in diffraction problems: **

**(1.46) **

Solving the equation **(1.45), the diffraction field outside the region of the object (in region V') is determined by integral transformation: **

**(1.47) **

The equations

**(1.48) **

**(1.49) **

where *r* = [(*x*–*x'*)² + (*y*–*y*')²]¹/², *H*0(2)(*x*) is the Henkel function of the second kind and the zeroth order, *G*(*x,y;x',y'*) *= i*/4 *H*0(2)(*kr*) is the Green function of the homogeneous space for the two-dimensional Helmholtz equation.

The equations **(1.48) and (1.49) solve the problem of diffraction of a cylindrical (two-dimensional) electromagnetic wave with TE-polarisation ( E0, E1 and E2 are the projections on the axis z of the vectors of the strength of the electrical field) on a cylindrical dielectric homogeneous object. Identical equations for TM-polarisation can be found in [3]. **

**1.3.2 Stratton–Chu equations **

Like the Green equations for the scalar functions **(1.37) and (1.38), the Green equations can be derived for the vector field. The starting point is the Gauss-Ostrogradskii formula: **

**(1.50) **

If the vector field **F **in the form of the vector product **F** = [**P**, rot **Q**] is substituted into equation **(1.50), we obtain the vector analogue of the second Green equation: **

**(1.51) **

*Taking into account the available vector relationships *

*Q rot rot P – P rot rot Q = PΔQ – QΔP + Q grad div P – P grad div Q = *

*= PΔQ – QΔP + div (Q div P – P div Q) *

equation **(1.51) has the new form: **

**(1.52) **

From the equations **(1.51) and (1.52) we obtain easily the integral relationships for the characteristics of the electromagnetic field in space. Let P = E, Q = aG, G(x – x0) = exp(ikR)/R, a is the unit vector of the arbitrary direction, where R = |x–x0|, x is the radius vector of the observation point, x0 is the point on the surface S. In this case, function Q satisfies the Helmholtz vector equation with the point source: **

**(1.53) **

where rot **Q** = [grad *G*, **a**], div **Q** = (**a**, grad *G*), *k*² = ω²εμ/*c*².

The vectors of the strength of the electrical and magnetic fields of a monochromatic lightwave in the homogeneous and isotropic space satisfy the heterogeneous Helmholtz equations:

**(1.54) **

where **j **is the density of auxiliary electrical current, ω is the cyclic frequency of vibrations, ε, μ is the dielectric and magnetic permittivity of the homogeneous medium, c is the speed of light in vacuum.

Using the relationships **(1.52)–(1.54) and the equation ( n, [E, [grad G, a]]) = (a, [grad G, [E, n]]), we obtain the equation (index 0 in div0, grad0, rot0 denotes differentiation with respect to x0): **

**(1.55) **

Taking into account that div **E** = 0, and the vector **a **is arbitrary, we obtain the integral representation:

**(1.56) **

where

By analogy with the equation **(1.56), we obtain the integral representation for the strength of the magnetic field: **

**(1.57) **

In the equations **(1.56) and (1.57), E0(x) and H0(x) are the strengths of the electrical and magnetic fields in the incident wave. **

The equations **(1.56) and (1.57) are referred to as the Stratton–Chu equations. **

*Diffraction on an object with ideal reflectivity *

*Diffraction on an object with ideal reflectivity*

As an example, we consider the solution of a problem of diffraction of an electromagnetic wave on an object with ideal reflectivity, occupying the region of space *V *with the surface *S*.

We introduce the notations for the surface densities of the electrical and magnetic current: **j***e*(**x**0) = [**n**,**H**(**x**0)], **j***m*(**x**0) = [**n**,**E**(**x**0)].

We take into account that grad *G *(**x**–**x**0) = –grad0 *G *(**x**–**x**0), rot (Φ, **F**) = Φ rot **F** + [grad Φ, **F**], rot *jm*(**x**0) = 0, rot **E** = *ik***H**

Consequently, instead of equation **(1.56) we obtain the following equation for the strength of the electrical field: **

**(1.58) **

Applying the rot operation to both parts of equality **(1.58) and taking into account that rot (grad Φ) = 0, we obtain the following equation for the strength of the magnetic field: **

**(1.59) **

by analogy, from equation **(1.57) we can obtain the integral representation for the strength of the electrical field: **

**(1.60) **

Taking into account the boundary conditions for the surface with ideal conductivity [**n**, **E**] = 0, (**n**, **H**) = 0, the equations **(1.59) and (1.60) can be presented in the new form: **

**(1.61) **

**(1.62) **

To derive the integral equation of the first kind for the density of electrical current on the surface of the ideal conductor, it is assumed that the vector **x **belongs to the surface. Multiplying the relationship **(1.61) by the vector of the normal at point x and taking into account the boundary condition for the ideal conductor, we obtain the integral equation: **

**(1.63) **

Thus, the problem of determination of the characteristics of the electromagnetic field is divided into two stages:

1) solution of the integral equation **(1.63) in relation to je (x0); **

2) calculation of the components of the field using the equations **(1.61) and (1.62). **

From the equation **(1.61) we obtain the integral Fredholm equation of the first kind with respect to the unknown current density of surface S through the known values of the strength of the magnetic field of the incident wave: **

**(1.64) **

*Diffraction on a transmission object *

*Diffraction on a transmission object*

We discuss the solution of the problem of diffraction of an electromagnetic monochromatic wave on a homogeneous dielectric object. For this purpose, we investigate the Maxwell equations inside the homogeneous region of the object *V1 *with the characteristics ε1 and μ and in the external region *V*2 with the characteristics of the medium ε2 and μ:

**(1.65) **

**(1.66) **

with the boundary conditions on the surface *S *of the interface of the media *V1 *and *V*2,

**(1.67) **

and with the radiation condition at infinity

**(1.68) **

Using the Green vector formula **(1.52), we obtain the integral Fredholm equation of the second kind for the vector of the strength of the magnetic field: **

**(1.69) **

The first term in equation **(1.69) can be regarded as the strength of the magnetic field in the known wave incident on the object: **

and the function of the pulsed response *G *(**x **– **x**0) satisfies equation **(1.53). **

After solving the equation **(1.69), the strength of the magnetic field in the external region V2 is determined using the integral transformation: **

**(1.70) **

The strengths of the electrical field **E**1 and **E**2 are determined from the known functions **H**1 and **H**2 from the Maxwell equations **(1.65) and (1.66). **

Instead of using the equations **(1.69) and (1.70) for determining the diffraction magnetic field, the Green vector formula (1.52) can be used to determine the Fredholm equation of the second kind for the determination of the strength of the electrical field in the diffraction region: **

**(1.71) **

where the known vector of the strength of the electrical field of the incident wave is expressed through the current density in the external region:

**(1.72) **

The vector of the strength of the electrical field of diffraction in the external region *V*2 is determined by solution of the equation **(1.71) and integral transformation: **

**(1.73) **

**1.4 Integral transforms in optics **

Using the scalar theory of diffraction, monochromatic light is described by the function of the complex amplitude *F*(**x**) = F(*x*, *y*, *z*) which satisfies the Helmholtz equation **(1.33): **

**(1.74) **

where *k *is the wavenumber. In the uniform and isotropic space without charges and current, the complex amplitude *F*(**x**) can be regarded as any projection of the vectors of the strength of the electrical **E**(**x**) and magnetic **H**(**x**) fields of the light wave.

Solving equation **(1.74) using the representation of the complex amplitude through the Fourier two-dimensional transform: **

**(1.75) **

where A (α, β, z) is the amplitude of the spatial spectrum of the plane waves, we obtain the expansion of the complex amplitude with respect to the plane waves:

**(1.76) **

where *A*0 (α, β) = *A *(α, β, z = 0) is also the amplitude of the spatial spectrum of the plane waves at *z* = 0. If we know the direction of propagation of light, then in the exponent in equation **(1.76) we retain only one sign (for the propagation of the wave along the axis z we selected the plus sign). **

Function *A*0 (α,β) is expressed through the Fourier inverse transform:

**(1.77) **

From the equations **(1.76) and (1.77) we obtain the integral transform of the complex amplitudes of the light field [4]: **

**(1.78) **

where

**(1.79) **

is the function of the pulsed response of the uniform space, *F*0 (*x*, *y*) = *F *(*x*, *y*, *z* = 0) is the complex amplitude of light at *z*appears below the integral and describes the nonuniform surface waves propagating in the plane *z* = 0 and providing no contribution to the light field at *z* λ. Therefore, under the condition *z* λ. the integral in equation **(1.79) can be calculated for α² + β² < 1 and not in infinite limits. **

**1.4.1 Kirchhoff integral **

Using the well-known expansion of the amplitude of the spherical wave with respect to plane waves

**(1.80) **

where *R* = (*x*² + *y*² + *z*²)¹/², the function of the pulsed response determined by the equation **(1.79) can be presented in the new form: **

**(1.81) **

If it is assumed that the distance from the plane *z =* 0 to the observation plane *z *is considerably greater than the wavelength *R* λ, *z* λ, instead of equation it can be approximately assumed that the following equality

**(1.82) **

is fulfilled. Consequently, instead of the integral transform **(1.78) we obtain the Kirchhoff integral: **

**(1.83) **

where *R* = [(*x *– *x*')² + (*y *– *y*')² + *z*²]¹/². In some cases, taking into account that *R *≈ *z, *instead of equation **(1.83) the Kirchhoff integral is written in the following form: **

**(1.84) **

The physical meaning of the Kirchhoff integral **(1.84) is linked with the Huygens–Fresnel principle, i.e., the Kirchhoff integral is the expansion of the complex amplitude of the light field with respect to the spherical waves. **

**1.4.2 Fresneltransform **

The integral transforms **(1.78) and (1.83) describe the distribution of non-paraxial light waves in an uniform space along the axis z. The propagation of paraxial light waves in a small solid angle is described by the Fresnel integral transform. **

The complex amplitude of the paraxial lightwave is presented in the form:

**(1.85) **

and the slowly changing complex amplitude *F*(*x*, *y*, *z*) satisfies the Fock-Leontovich parabolic equation **(1.34) [2]: **

**(1.86) **

is the transverse Laplacian. Any solution of the equation **(1.86) can be presented in the integral form: **

**(1.87) **

Where *F*0 (*x*,*y*,*z*) = *F *(*x*,*y*,*z* = 0).

The Fresnel transform **(1.37) is the expansion of the complex amplitude of the paraxial light wave with respect to the parabolic waves, and can be easily obtained from the Kirchhoff integral (1.84) using expansion into a Taylor series to the second term of distance R in the exponent: **

The transition from the Kirchhoff integral **(1.84) to the Fresnel integral (1.87) is possible when the following condition **

is satisfied, where *r *is the effective radius of the light field.

At a large distance from the initial plane *z* = 0 when the conditions of the far diffraction zone (or the Fraunhofer diffraction zone) are fulfilled:

**(1.88) **

the Fresnel integral transform **(1.87) can be replaced by the Fourier transform with the multiplier of the parabolic wave in front of the integral: **

**(1.89) **

The Fourier transform **(1.89) is the expansion of the complex amplitude of the paraxial lightwave with respect to plane waves. The integral in the right-hand part of equation (1.89), written using the spatial frequencies u = kξ/z, v = kη/z, has the form of the conventional Fourier transform: **

**(1.90) **

The Fourier transforms **(1.89) and (1.90) also describe the complex amplitude of the light field in the plane of the spatial frequencies of a thin spherical lens. **

**1.5 Methods for solving the direct diffraction problem **

**1.5.1 Different schemes of solving differential equations **

We formulate the initial problem as follows: it is required to find function *u *which satisfies in all internal points of the region ω the equation (or a system of equations) in partial derivatives, and in the sections of the boundary Γ – the essential conditions ensuring that the solution of the problem is correct (the solution exists, uniquely and stably in relation to the small perturbations of the input data). This problem will be denoted as follows:

**(1.91) **

where *L *is the operator which determines the type of differential equation (system) and boundary conditions, *f *are the appropriate right-hand parts.

The concept of the different method is based on replacing the differential problem **(1.91) by the difference scheme [5]: **

**(1.92) **

where operator *Lh *is determined by the substitution of the derivatives in the operator *L *by finite differences, and functions *uh *and *fh *are determined in the grid region ω*h *consisting of a finite set of the points of the region ω situated at the characteristic distance *h *from the selected adjacent points. Further, the solution of the **problem (1.92) is reduced to solving the system of algebraic equations with respect to the grid function uh. **

It is assumed that *uh *corresponds with some accuracy to the solution *u *of the initial **problem (1.91). This situation exists if the condition of convergence of the solution of (1.92) to the solution of (1.91) is fulfilled: **

**(1.93) **

where [u]*h *are the values of the function *u *in the region ω*h *and ||..||*Uh *is the norm in the space of the grid functions *Uh*.

**(1.94) **

Since the finite differences correspond to the derivatives with some error, substitution of the solution of **(1.91) into the scheme (1.92) results unavoidably in discrepancy δ fh: **

The difference scheme approximate the initial differential problem if

where ||*..*||*Fh *is the norm in the space of the mesh functions *F*h. In addition, when the condition

is fulfilled, the approximation with the order O (*hk*) applies.

Schemes stable with respect to small perturbations of the initial data are used in practice.

The difference scheme is assumed to be stable if there are constants *h*0 and δ0 such that at any *h < h*0 and any grid function ε*h*, ∈ *Fh *such that

the difference problem

obtained from **(1.92) by adding the perturbation ε h to the right-hand part, has the unique solution, and the following estimate holds **

The convergence theorem [**5] links the convergence of (1.93) with the concepts of approximation and stability. Let the difference scheme (1.92) approximate the problem (1.91) with the order O ( hk) and let the difference scheme be stable. Consequently, the solution of the difference problem is reduced to solving the differential problem, and the estimate (1.94) holds. In this case, both convergence and stability should be proven in agreed norms. **

**1.5.2 The finite element method for solving integral equations **

The finite element method (FEM) is the method of approximate solution of the differential or integral equations of mathematical physics using the expansion of the sought functions with respect to the approximate orthogonal bases.

Three most frequently encountered methods of approximate solution of linear operator equations will be examined. It is the Ritz variance methods, the method of least squares, and the Galerkin projection method [**5]. **

The operator form of the heterogeneous linear differential equation is:

**(1.95) **

where, for example, the operator for the Helmholtz equation has the form:

*u *is the sought function of the complex amplitudes, *f *is the source function. We introduce the scalar product of two functions:

**(1.96) **

*The Ritz method *

*The Ritz method*

Using the Ritz variance methods [**5], the approximate solution u(n) of equation (1.95) is found in the form of expansion with respect to the known set of the basic (linearly independent) functions ψ(n): **

**(1.97) **

and the unknown expansion coefficients *Cm *are determined from the condition of minimum of the functional:

**(1.98) **

The minimum of the functional **(1.98) is obtained with the functions which are solutions of Euler equations: **

**(1.99) **

the system of linear algebraic equations **(1.99) with the expansion (1.97) and (1.19) taken into account, can be written in the matrix form: **

**(1.100) **

where the elements of the Gram matrix have the form:

**(1.101) **

and the column vectors consist of the unknown coefficients and projections of the function of sources on the basic functions:

**(1.102) **

The basic functions ψ*i*(*n*) in the finite element method are the linear (quadratic or of a higher order) approximations of the function of the complex amplitude of the field in the *i*-th cell of the space, if the entire region of the field *V *is divided into *Nn *non-intersecting sub-regions.

*The method of least squares *

*The method of least squares*

The method of least squares [**6] is also a variance method which can be used for the approximate solution of equation (1.95) by minimising the functional of the type: **

**(1.103) **

The functional **(1.103) has a minimum on the functions which are the solution of the system of Euler equations (1.99). The sought function of the complex amplitude is presented in the form of the linear combination of the known approximating linearly independent basic functions with the unknown coefficients (1.97) which are found from the system of linear algebraic equations (1.100). The matrix of the system (1.100) and the column vectors have the following form in this case: **

**(1.104) **

is symmetric and positively determined, and the sequence of approximate solutions *u*(n) from equation **(1.97) is reduced in the norm to the exact solution of equation (1.95) at n → ∞. **

*The Galerkin method *

*The Galerkin method*

The Ritz method can be used for the equations **(1.95) with self-conjugate and positively determined operator [5]. However, the operator of the Helmholtz equation (1.74) is not positively determined. The equation can be solved by the finite element method using the Galerkin projection method (it is sometimes referred to as the Bubnov–Galerkin method). **

from equation **is the symmetric and positively determined operator, then the unknown coefficients Ci in the expansion of the function of the approximate solution u(n) with respect to the basis of the finite elements ψm(n) (1.97) are determined by solving the system of linear algebraic equations written in the matrix form, as in the case of equation (1.100). The elements of the matrix and the vector column in the right-hand part of equation (1.100) have the form: **

**(1.105) **

If equation **(1.94) has the unique solution and operator L^‒ 1KT is fully continuous, the approximate solutions u(n**

**Conclusion **

This chapter deals with the main differential and integral equations which are essential for solving the problems of diffraction of electromagnetic waves. Using the general system of differential equations for the vectors of the strength of the electrical and magnetic fields of the electromagnetic wave, the wave equation, the Helmholtz equation for the monochromatic waves, the Fock–Leontovich equation for the paraxial light fields, and also the eikonal equation, describing the propagation of beams in geometrical optics were gradually derived. The scalar and Green vector theorems were used to derive the main integral relationships for the monochromatic electromagnetic field: the Stratton–Chu and Kirchhoff–Helmholtz equations. The main Fredholm integral equations of the first and second kind were represented for solving the problem of diffraction of monochromatic electromagnetic waves on homogeneous dielectric (transmitting) objects with ideal reflectivity. The frequently used integral representations: the expansion of the amplitude with respect to flat waces, expansion with respect to spherical waves (Kirchhoff integral), expansion with respect to parabolic waves (Fresnel transform), were investigated for the scalar complex amplitude which can be represented by any of the projections of the vectors of the strength of the electrical and magnetic fields.

Many of the relationships presented in this chapter will be used in the chapters of this book for solving direct and inverse problems of diffractive computer optics.

**References **

1. Born, M., Wolf, E. *Principles of optics*. Oxford: Pergamon Press; 1973.

2. Solimeno, S., Crosignani, B., di Porto, P.Diffraction and waveguide propagation of optical radiation [Russian translation]. Moscow: Mir, 1989.

3. Il’inskii, A.S., Kravtsov, V.V., Sveshkinov, A.G.Mathematical models of electrodynamics. Moscow: Vysshaya shkola, 1991.

4. Zverev, V.V. Radio-optics, Sov. radio, 1975. [Moscow].

5. Ortega, J.M., Poole, W.G. *An introduction to numerical methods for differential equations*. Marshfield, MA: Pitman; 1981.

6. Marchuk, G.I., Agoshkov, V.I. *Introduction to design-network methods*. Moscow: Nauka; 1981.

7. Mikhlin, S.G. *Variance methods in mathematical physics*. Moscow: Nauka; 1970.

**2 **

**Diffractive optical transformations **

**2.1 Transformations in optical systems **

As indicated by the material in the first chapter, the diffraction equations are linear, i.e., the superposition principle is fulfilled for these equations. Considerable advances have been made in optics using the approach developed in the mathematical theory of linear systems with the distributed parameters [**1]. The important concepts will now be described. **

**Optical signal **

To simplify considerations, it is assumed that the light is linearly polarised. The electrical field at the moment of time *t *at the point with the coordinates **x** = (*x*1, *x*2, *x*3), excited by a monochromatic light source, can be written in the complex form:

**(2.1) **

is the frequency of the light source, λ is the wavelength, *c *is the speed of light,

**(2.2) **

is the optical signal with the amplitude *A *(**x**) and phase φ(**x**).

Equation **(2.1) in which the spatial and time variables are separated, can be used also for a quasi-monochromatic light source in which the width of the frequency band Δ ω is considerably smaller than the mean frequency of emitted light: **

**(2.3) **

*Example 2.1 *

*Example 2.1*

A spherical wave is described by the equation

**(2.4) **

The surface of the constant phase is a plane.

It should be mentioned that a spherical lens transforms the spherical wave into a plane wave and vice versa, as shown in **Fig. 2.1 **

**Fig. 2.1 **Transformation of a spherical wave to a plane wave.

**(2.5) **

**Two-dimensional linear systems **

The concept of the optical system transforming images in accordance with the rules determined by the set of the optical elements incorporated in the system, and by the mutual relationship between these elements, is well-known from physics lectures.

From the mathematical viewpoint, the system is the rule L setting the output function g as corresponding to the input function f. There are one-dimensional (1-D) and two-dimensional (2-D) systems. The one-dimensional systems transform the functions of a single variable:

**(2.6) **

Correspondingly, the two-dimensional systems transform the functions of two variables:

**(2.7) **

Because of their nature, the optical systems are two-dimensional but in some cases they can be treated as one-dimensional.

A special position in the group of all possible systems is occupied by the linear systems. A system is linear if it is governed by the superposition principle according to which the response of a system to the weighted sum of two input effects is equal to the weighted sum of the responses to each effect, i.e.

**(2.8) **

where *a*1, *a*2 are the weight coefficients which can be represented by complex quantities.

The superposition principle can be expressed in a more general form, examining an arbitrary number *K *of input effects

**(2.9) **

The concept of the point light source plays a fundamental role in studies of optical systems. The point source is characterised by an infinitely high density of brightness probabilities in an infinity small spatial domain, in a point:

**(2.10) **

The delta function, which is located at an arbitrary point with the coordinates (ξ1, ξ 2), and not at the origin of the coordinates, can be described by the equation:

**(2.11) **

We examine a two-dimensional linear system whose input receives a signal in the form of a delta function. The reaction of the system to the delta function differs for different systems. It is referred to as the pulsed response and is characteristic of the 2-D system:

**(2.12) **

Using the function of the pulsed source, we can write an equation linking the images at the input and output of the 2-D linear optical system:

**(2.13) **

The system is referred to as spatially invariant if its pulsed response depends on the difference of the coordinates of the input (*x*1, *x*2) and output (ζ1, ζ2) planes. For the optical system shown in **Fig. 2.2, this means that with the movement of the point source in the input (object) region, the image of this object in the observation plane changes the positions but not the shape. **

**Fig. 2.2 **Optical spatially-invariant system.

For the spatially-invariant systems, the pulsed response is described by the function:

**(2.14) **

Instead of equation **(2.13) we obtain: **

**(2.15) **

Equation **(2.15) is the convolution integral. **

When describing the spatial-invariant optical systems, the pulsed response is sometimes replaced by the frequency characteristic, determined by the two-dimensional Fourier transform:

**(2.16) **

Or in the symbolic form:

**(2.17) **

where *Φ *- is the Fourier transform symbol.

The relationship between the spatial spectrum of the input and output optical signals is described by the algebraic equation:

**(2.18) **

Equation **(2.18) is the initial equation for optical spatial filtration. **

Here, we can discuss the transformations of the signals, carried out in the optical systems or ‘optical transformations’.

**The element base of optical transformations **

The classic element base of optical systems – lenses, prisms, mirrors, diaphragms – has been supplemented in recent years by holographic optical elements, including digital holograms and diffractive optical elements (DOE) with extensive functional possibilities, synthesised using computers. The diffractive optical elements include flat lenses, programming diffraction gratings, laser beam dividers, aberration correctors, focusing devices, modulated diffractive gratings, special filters, beam shapers and many other optical signal converters, based on the efficient application of the diffraction phenomenon. Strictly speaking, it is not completely correct to refer to this group of devices as ‘optical elements’ because these elements often play the role of several ‘classic’ optical elements. For example, a focusator into the coaxial section has the function of a spherical lens and an axicone, and a focusaor into the transverse section replaces two crossed cylindrical lenses and one spherical lens. In addition, it is often difficult to use the name optical element for holographic spatial filters or modulated diffractive gratings. However, the term ‘diffractive optical element’ is now used widely and we shall also use it, specifying in important cases the meaning of the optical transformation performed by this element. It is also necessary to clarify the widely used term ‘diffractive optics’. From the viewpoint of dividing science into individual sections, it is a section of optics concerned with the investigations of diffractive phenomena. Diffractive optics as a section of optical technology has recently been referring to the entire range of diffractive optical elements. There is no contradiction between these two interpretations, because the situation in the case of the term ‘optics’ is on the whole the same.

**Fourier optical transform **

We study the problem of diffraction of an optical signal on an orifice S in a non-transparent screen, placed in the plane *z* = 0 (**Fig. 2.3). **

**Fig. 2.3 **Diffraction on an orifice.

The field *g *(*x*, *y*) at point *P *(*x*, *y*) positioned at a large distance *z*0 is expressed in the form [**1]: **

**(2.19) **

where

**(2.20) **

is the transformation of the Fourier function *f*(*u*, *v*); α = *x*/*r *and β = *y*/*r *are the directing cosines of the OP line; A is the slowly changing coefficient; *k* = 2π/λ is the wave number.

Thus, the investigated optical system carried out, with the accuracy to the phase coefficient, the 2-D Fourier transform of the input signal with respect to spatial variables. The investigated solution of the diffraction problem is referred to as the Fraunhofer diffraction.

It should be mentioned that if the diaphragm *S *is replaced by a thin spherical lens, the rear focal plane of the lens also shows the distribution of the field proportional to the 2-D Fourier transform of the input signal.

If the orifice *S *is covered with a thin transparent material with the transmission function *T *(*u*, *v*), the field in the remote zone is described by the equations **(2.19), (2.20) on the condition that in equation (2.19) function f (u, v) is replaced by the product f(u, v) T(u, v). Selecting different functions T (u, v), we obtain the appropriate diffraction patterns. **

*Example 2.3 *

*Example 2.3*

Let the incident wave be homogeneous, i.e. *f *(*u*, *v*) = 1, and the orifice is rectangular with the sides 2*a *and 2*b *(**Fig. 2.4). **

**Fig. 2.4 **Example of a rectangular orifice.

The Fourier transform, describing the diffraction pattern, has the form:

**(2.21) **

*Example 2.4 *

*Example 2.4*

Let us assume that a non-transparent screen has an 1-D grating with 2 (*M* + 1) orifices, formed by multiple shift by period *d *of the function shown in **Fig. 2.4 along the axis u (Fig. 2.5). **

**Fig. 2.5 **Example of the 1-D grating.

The expression for the spectrum has the form [**3]: **

**(2.22) **

In the remote zone, the quantity *x*/*z*as the reduced angular coordinate (**Fig. 2.6). **

**Fig. 2.6 **Diffraction on the amplitude binary grating

Correspondingly, we investigate the distribution of intensity of the diffraction field with respect to the angles:

**(2.23) **

**Figures 2.7 a and 2.7b show the graphs of the normalised function |F(u,0)|²|u = ωa for the values of the parameters d = 4a, M = 5 and 10. **

**Fig. 2.7 **(a, b) Distribution of intensity of the diffraction field with respect to angles.

**2.2 Diffraction gratings **

A diffraction grating is a set of a large number of obstacles and orifices concentrated in a limited space on which light is diffracted [**2]. There are non-regular and regular diffraction gratings. A diffraction grating is non-regular if the orifices and the obstacles are distributed at random. In a regular grating, the orifices and obstacles are distributed in accordance with a specific law. The diffraction gratings are of the surface and spatial type. Regular surface diffraction gratings are important for practice. They will now be discussed. **

**One-dimensional diffraction gratings **

The first DOEs – diffraction gratings – were developed more than two centuries ago. The amplitude one-dimensional diffraction grating is a flat transparent on which dark and light parallel bands (strips and slits) alternate. The bands have the same width. **Figure 2.8 shows a fragment of one-dimensional amplitude diffraction gratings, and Fig. 2.9 is the transmission function of light along the grating in the direction of the x axis. **

**Fig. 2.8 **Fragment of a binary amplitude diffraction grating.

**Fig. 2.9 **Transmission function of the amplitude diffraction grating.

If such a grating is illuminated with a monochromatic light beam with the wavelength λ incident normal to the plane of the grating, diffraction of light on the periodic structure of the slits leads to the formation of a large number of light beams exiting under different angles α*p *and corresponding to different diffraction orders. Angle α*p *depends on the period of the grating and, assuming that the angles are small, is determined from the equation:

**(2.24) **

where *p* = 0, ± 1, ± 2,…, *d *is the grating period.

The intensity of light decreases with increasing absolute value *p*. As indicated by the theory of diffraction gratings, for the same width of the strips and the slits, the nature of this decrease is determined by the equation:

**(2.25) **

where *a* = π*d*/2λ. Since a large part of light is absorbed by the amplitude transparant, the diffraction efficiency in the first order does not exceed 10%. In the one-dimensional (1-D) phase diffraction grating this parameter is approximately 4 times higher. A fragment of a one-dimensional phase diffraction grating is shown schematically in **Fig. 2.10. **

**Fig. 2.10 **Fragment of a binary phase diffraction grating.

In this case, the transmission function is purely a phase function, periodic with period d and changing abruptly by π (**Fig. 2.11). From the functional viewpoint, the phase and amplitude diffraction gratings are identical. **

**Fig. 2.11 **Phase of the transmission function of the phase diffraction grating.

The ‘basic’ diffractive optical elements, such as 1-D diffraction gratings, can be improved by changing the zone profile. This yields the so-called blazed gratings. For example, we can examine the amplitude 1-D diffraction grating, whose transmission function changes in accordance with the law:

In theory, in light diffraction on this grating, only ± 1 diffraction orders form in addition to the zero order. In practice, taking into account the errors in manufacture of the DOE, the situation is far more complicated.

**Zone plates **

A natural development of one-dimensional binary amplitude and phase diffraction gratings are the appropriate two-dimensional (2-D) radial-symmetric gratings (zone plates).

**Figure 2.12 shows the central fragment of a Rayleigh–Soreau zone plate, and Fig. 2.13 is the transmission function in relation to radius. The zone plate is used for focusing light and behaves as a lens with a large number of focuses, f–2, f–1, f0, f1, f2, corresponding to different diffraction orders and distributed on the light propagation axis: **

**Fig. 2.12 **Central fragment of a Rayleigh–Soreau zone plate.

**Fig. 2.13 **The transmission function of the amplitude zone of the plate.

**(2.26) **

The normalised light intensity decreases with the increase of the order number in accordance with the law:

**(2.27) **

The concept of the zone will be explained. The zone is a restricted area of the DOE on which the transmission function of light undergoes a single variation from the minimum to maximum value. For an amplitude diffraction grating, the zone is a combination of dark and light bands within the limits of the single period of the grating (**Fig. 2.8). In this case, the boundary of the zones is a straight line. For the Rayleigh–Soreau zone plate [1], the zone consists of a set of dark and light rings of variable thickness (Fig. 2.12). **

In this case, the boundary of the zones is a circle with a variable diameter. Further, we consider diffractive optical elements with more complicated zones and, correspondingly, the zone boundaries.

The radii of the circle change in proportion to the square root of the consecutive integers *p*:

**(2.28) **

where *f *is the focusing distance.

The fragment of the 2-D phase zone plate and the appropriate phase of the transmission function are shown in **Fig. 2.14 and 2.15. **

**Fig. 2.14 **A fragment of a binary phase zone plate.

**Fig. 2.15 **Phase of the transmission function of a binary phase zone plate

*Two-dimensional diffraction gratings *

*Two-dimensional diffraction gratings*

The two-dimensional phase gratings can be produced from one-dimensional ones. For this purpose, it is necessary to multiply the phase functions of two one-dimensional gratings. From these two one-dimensional binary gratings we obtain a two-dimensional binary grating. The diffraction efficiency of such a grating is lower. Thus, if we use a one-dimensional grating with 66% efficiency, the efficiency of the produced two-dimensional grating cannot be higher than 44%.

**Figure 2.16 a shows an example of a two-dimensional grating, generating nine identical orders. The grating was produced from two identical one-dimensional phase gratings and, consequently, is symmetric. For this reason, the grating generates the symmetric distribution of the intensities. This method can be used to produce different phase gratings. They can generate different two-dimensional distributions in relation to the initial gratings. **

**Fig. 2.16 **Phase profile of a two-dimensional nine-order binary grating (fragment) (a). Distribution in the Fourier plane, diffraction efficiency 41.25% (b).

The same procedure was used for producing a phase grating for 49 orders of the same intensity (**Fig. 2.17 a, 2.17b). **

**Fig. 2.17 **Phase profile of a two-dimensional 49-order binary grating (fragment) (a). The distribution in the Fourier plane, efficiency 54.33% (b).

**2.3 Flat lenses and prisms **

**Reducing the function to the interval **

The phase function φ(*u*, *v*) of optical element is a function describing the dependence of the phase incursion, produced by the optical element at the point (*u*, *v*) on the spatial coordinates. In a number of cases, the boundaries of the zones of the diffractive optical elements can be determined on the basis of the phase functions of the appropriate traditional (refraction) optical elements.

The traditional optical element is characterised by a smooth phase function φ(**u**), where **u** = (*u*, *v*) are the transverse coordinates of the point situated in the plane of the optical element. In illuminating an optical element with a beam with complex amplitude *W*0 (**u**) a field with the amplitude

**(2.29) **

forms immediately behind the optical element. **Figure 2.18 shows the phase function of the spherical lens. **

**Fig. 2.18 **Phase function of the spherical lens.

The thickness of the traditional optical element *H *may equal tens of thousands of wavelengths. The values of the phase φ are distributed in the range from 0 to thousands of units. At the same time, the variation of the phase in the previously examined DOEs are found in the range from 0 to 2π.

Taking into account the fact that the complex exponent in equation **(2.29) is a periodic function with a period 2π, phase can be reduced to the interval [0, 2 π], **

**(2.30) **

where mod2*π*φ = φ − 2π*j* at *j*2π ≤ φ ≤ (*j* + 1)2π, *j* = 0, ± 1, ± 2, …

For practical considerations it may prove to be useful to reduce the phase to the interval [0, 2π*m*) using the equations

**(2.31) **

where usually m ~1–10². The equivalence of the initial smooth phase φ and of the piecewise phase *Φ *reduced to the interval is expressed by the relationship:

**(2.32) **

**A flat spherical lens **

The equation of the phase function of the spherical lens in the paraxial approximation has the form:

**(2.33) **

is the wave number, *f *.

The reduction of the phase function **(2.33) to the interval [0, 2π] is shown in Fig. 2.19. If the material of the lens has the refractive index n, the maximum height of the relief is: **

**Fig. 2.19 **Reduction of the phase function of the lens to the interval [0, 2π).

and has the order of the wavelength.

The radii of the Fresnel zones can be determined from the relationship:

**(2.34) **

from which it follows that

**(2.35) **

The number of the total zones *F *and satisfies the relationship

**(2.36) **

where]∙[denotes the integral part of the number with rounding to the smaller side.

is variable and decreases in the direction to the lens periphery. The characteristic parameter is the width of the narrowest (the last in the present case) peripheral zone which determines the requirements on technological equipment. In a more general case of a DOE with the phase function φ(*u*, *v*) we can determine the width of the narrowest zone of the microrelief with the maximum phase height 2π*m*. Using the linear approximation of the phase function within the limits of this zone, we obtain the equation:

**(2.37) **

is the gradient of the function φ(*u*, *v*).

We compare the equations of the radii of the rings of the Rayleigh–Soreau zone **plate (2.21) and the spherical flat lens (2.35). It may be seen that the radii of the zones on the lens are √2 greater than the radii of the rings on the zone plate. Correspondingly, for the given diameter, the number of the rings on the zone plate is twice as high, i.e., every ring of the spherical flat lens is related to two rings of the zone plate: dark and light. The flat lens has a continuous (within the limits of the zone) phase relief and the phase zone plate has the binary phase relief which approximates the former (see Fig. 2.20). **

**Fig. 2.20 **Phase function of the Fresnel lens and the binary zone plate (cross-hatched regions).

If the formation of the diffractive optical elements is based on discrete approximation of the continuous phase function, we can use the procedure of stepped