GB2440164A

GB2440164A - Apparatus for and a method of determining surface characteristics

Info

Publication number: GB2440164A
Application number: GB0614358A
Authority: GB
Inventors: Daniel Ian Mansfield
Original assignee: Taylor Hobson Ltd
Current assignee: Taylor Hobson Ltd
Priority date: 2006-07-19
Filing date: 2006-07-19
Publication date: 2008-01-23
Also published as: GB0614358D0

Abstract

Light (4) is directed along a sample path (SP) towards a sample surface (7) and along a reference path (RP) towards a reference surface (6) such that light reflected by the surfaces interfere. A mover (11) effects relative movement along a scan path between the sample surface (7) and the reference surface (6). A detector (10) provides a series of intensity values representing interference fringes produced by a sample surface. A processor (32) receives first intensity data resulting from a measurement operation on a surface area of a substrate and second intensity data resulting from a measurement operation on a surface area of a thin film structure. The processor (32) determines a gain (100) for the film structure(s) and determines a substrate surface characteristic (101) on the basis of the first intensity data, determines an apparent thin film structure surface characteristic on the basis of the second intensity data, and modifies the apparent thin film structure surface characteristic using the substrate surface characteristic and the gain(s) determined by the gain determiner.

Description

APPARATUS FOR AND A METHOD OF DETERMINING

SURFACE CHARACI'ERISTICS This invention relates to apparatus for and a method of determining surface characteristics such as surface roughness of a thin film structure using interferometric techniques.

Techniques such as phase-stepping interferometry(PSI) and coherence scanning or broadband scanning interferometry (sometimes called "white light scanning interferometry" or "scanning white light interferometry" (SWLI)) have been used to determine surface topography. Phase shifting interferornetry involves the acquisition of a number, generally four, of interferograms, at predetermined, usually quarter wavelength, intervals and the processing of these interferograms to determine the surface height of a region (surface pixel) of a surface under examination. The phase shifting method is explained in sections 14. 4 and 14.5 on pages 506 to 515 of "Optical Shop Testing", Second Edition by Daniel Malacara (ISBN 0-471-52232- 5). Coherence scanning or broadband interferometry determines the surface height ofa surface pixel by acquiring interferograms as the path difference between the reference and measurement arms of the interferometer is changed and identifying the interferogram, and thus the position along the scan path, at which the coherence peak occurs for that surface pixel as discussed in a paper entitled "Profilometry with a Coherence Scanning Microscope"by Byron S. Lee and Timothy C Strand published in Applied OpticsVolume 29, No. 26 10 September 1990 at pages 3784 to 3788. Coherence scanning or broadband scanning interferometry has the advantage of a longer range than phase shifting interferornetry.

Measurement of the surface roughness of thin fllmstructures by PSI or coherence scanning has proved difficult because the topography-induced phase change is corrupted by the presence of the thin film(s). Thus, in the case of coherence scanning or broadband scanning interferometry, it has proved possible to measure the surface roughness only where the layers have a thickness exceeding the coherence length of the interferometer light source so that there is a coherence peak associated with every interface and topographical perturbation measurements can be made of the film interfaces. As is well-known, the coherence length of

S r

an interferometer is a function of bandwidth and numerical aperture (NA).

In one aspect) the present invention provides apparatus for and a method of determining the surface roughness of a thin film structure by modifying a measured apparent surface roughness in accordance with a model representing the difference or apparent gain in the measured surface height of a surface pixel of a layer compared with the actual surface height of that surface pixel where the gain model takes into account the effect of phase change on reflection at interfaces of the thin film structure and instrument profile characteristics.

In operation of an embodiment, light from a light source is directed along a sample path towards a region of a sample surface and along a reference path towards a reference surface such that light reflected by the region of the sample surface and light reflected by the reference surface interfere. A mover effects relative movement along a scan path between the sample surface and the reference surface. A detector senses light intensity at intervals to provide a series of intensity values representing interference fringes produced by a region of a sample surface. A data processor receives first intensity data comprising a first series of intensity values resulting from a measurement operation on a surface area of a substrate and second intensity data comprising at least a second series of intensity values resulting from a measurement operation on a surface area ofa thin film structure. The data processor has a gain determiner that determines a gain for the or each thin film of a thin film structure and a surface characteristic determiner that determines a substrate surface characteristic on the basis of the first intensity data, that determines an apparent thin film structure surface characteristic on the basis of the second intensity data, and that modifies the apparent thin film structure surface characteristic using the substrate surface characteristic and the gain or gains determined by the gain determiner.

As used herein the term "gain" means the rate of change of the apparent surface height with thin film thickness.

Embodiments of the present invention will now be described, by way of example, with reference to the accompanying drawings, in which: ( Figure 1 shows a schematic block diagram of apparatus for determining a surface characteristic such as surface roughness of a thin film layer structure, the apparatus comprising an interferometer system and a data processing and control apparatus; Figure 2 shows a functional block diagram of computing apparatus that may be configured to provide the data processing and control apparatus shown in Figure 1; Figure 3 shows a flow chart to illustrate operations carried out by a gain determiner of the data processor shown in Figure 1; Figure 4 shows a flow chart to ifiustrate operations carried out by a surface characteristic determiner of the data processor shown in Figure 1; FigureS shows a graph of the gain determined by the gain determiner against optical thickness for single thin films on substrates where thesingle thin films have different refractive indices and the interferometer is a monochromatic low numerical aperture interferometer; -Figure 6 shows a graph of reflectance r against film thickness for a range of different refractive indices; Figure 7 shows a graph of the gain determined by the gain determiner against optical thickness for a silicon film on a silicon dioxide substrate where the interferometer is a 450 nm wavelength interferometer with a x50 magnification Mirau objective; Figure 8 shows part of the graph show in Figure 7 on an enlarged scale; Figure 9 shows a functional block diagram of apparatus embodying the invention in greater detail than shown in Figure 1 and where the apparatus is configured to carry out coherence scanning interferometry; p Figure 10 shows a simplified side-elevational, part sectional view of one example of a suitable interferometer system for use in the apparatus shown in Figure 9; Figure 11 shows a graph of intensity I against position Z to illustrate an example of the interference fringes (axial interferogram) produced by coherence scanning interferometry for a sample surface region around a coherence peak or interference region; Figure 12 shows a graph of the gain determined by the gain determiner against optical thickness for a silicon film on a silicon dioxide substrate where the apparatus is configured to carry out coherence scanning interferometry and has a xSO magnification Mirau objective; and Figure 13 shows a simplified block diagram of a thin film structure manufacturing system.

Referring now to the drawings, Figure 1 shows a simplified schematic block diagram of apparatus 1 for determining surface characteristics, in particular the surface roughness ofa thin filmstructure consisting of a substrate or base carrying one thin film layer or a stack of two or more thin film layers, where the surface roughness is the root mean square surface deviation corresponding to spatial wavelengths in excess 01' approximately I micrometre. A definition of surface roughness is to be found in ANSI standard B46. 1 2002. As used herein, thin film layerineans a substantially transmissive film layer that is sufficiently thin to exhibit interference effects, in practice a film layer having a thickness in the region of lOs of nano metres to lOOs of nanometres so that the mean thickness of the film is less than the coherence length of the light source where the interferometer is a focussed interferometer.

The apparatus I has an interferometer system 2 and data processing and control apparatus 3.

The interferometer system 2 is based on a conventional inter1rometer and may have a Mirau, Michelson or Linnik configuration, for example. Where the interferometer system is configured to carry out phase shifting interferometry, then the interferometer will have a monochromatic light source 4 whereas where the interferometer system 2 is configured to carry out coherence scanning interferometry, the interferometer will have a broadband source 4 which may be, for example, a white light source such as a quartz halogen lamp coupled to a regulated DC power supply having a light intensity output user control 400 provided, for example, in the form of a user-rotatable knob.

The source 4 provides light L which is split by a beam splitter (illustrated in Figure 1 as a single beam splitting prism) 5 into a first light beam which is directed along a reference path RP towards a reference mirror 6 and a second light beam which is directed along a sample path SP towards a surface 7 of a thin film structure 8 mounted on a sample support stage 9. Light reflected from the reference mirror 6 returns along the reference path RP to the beam splitter 5 where it interferes with light reflected from the surface 7 back along the sample path SP. A focussing element 3 may be provided to focus an image of the region of interference onto a detector 10.

In this embodiment, the detector 10 has a 2D (two-dimensional) array SA of image sensing elements SE, one row of which is shown very diagrammatically in Figure 1. The array SA images an area of the sample surface 7 falling within the field of view of the detector 10. Each individualsensing element SE of the 2D sensing array of the detector 10 detects the portion of the interference pattern falling within the acceptance cone of that element and resulting from a corresponding surface region or surface pixel of the area of the sample surface 7 so that, effectively, theimaged area of Lhe surface can be considered as a 2D array of surface regions or surface pixels. In this example, the detector 10 isa CCD (Charge Coupled Device) digital camera, for example, a Vosskühler GmbH: CCD 1300 CCD digital camera. As another possibility, a CMOS detector having a 2D (two-dimensional) xy array of CMOS sensing elements may be used. In either case, generally each of the sensing elements is square to provide the same resolution in both directions (x and y) of the array.

A motion controller 11 is provided to effect relative movement between the thin film structure and the reference mirror 6 so as to change the difference between the lengths of the paths travelled by light reflected from the reference mirror 6 and light reflected from the sample surface 7. As shown in Figure 1, the motion controller ii is arranged to move the reference mirror 6 along the reference path RP. This is equivalent to moving the sample surface 7 along r a scan path in the Z direction shown in Figure 1.

The detector 10 is arranged to capture or sense the light intensity (i.e. the interference pattern) at intervals as the path length difference between the thin film structure and the reference S mirror 6 is changed. Where the interferoineter system is configured to carry out phase shifting interferometry, then the motion controller will cause a number, generally four, of A/4 phase steps where is the wavelength of the monochromatic light source and the detector will capture or sense the light intensity at each phase step. Where the interferometer system is configured as a coherence scanning interferometer system the motion controller will cause relative movement along a scan path and the detector will capture or sense the light intensity at intervals corresponding to axial changes in the relative position of the thin film structure of, for example, A/8, where A is the nominalcentral wavelength of the broadband source and the step may be for example 7Snm. 2D image or frame data representing the intensity pattern for the field of view of the detector 10 is acquired by the detector 10 at each interval.

As shown in Figure 1, the data processing and control apparatus 3 has control apparatus 30 for controlling operation of the interferometer system 2, an intensity data receiver 33 for receiving intensity data signals from the detector 10, a data processor 32 for processing the intensity data under the control of a controller 21 of the control apparatus 30 and a user interface 3lfor enabling a user or operator: to control operation of apparatus; to be provided with a data output representing the results of processing by the data processor 32 of the data acquired during a measurement operation; and to enable messages such as error messages to be communicated to the user.

The controller 21 of the control apparatus 30 controls overall operation of the apparatus and communicates with the user interface 31 and data processor 32.

The data processor 32 is operable to determine surface characteristics such as surface roughness ofa thin film structure and to this end comprises a surface characteristic determiner 101 for determining the interfacial surface roughness and a gain determiner 100 for determining a gain value or values for use by the surface characteristic determiner 101 in determining the r interfacial surface roughness, where the term "gain" means the rate of change of the apparent surface height with thin film thickness and represents the effect on the surface roughness measurements of the actual thin film structure being measured (for example the effect of refractive indices and film thicknesses) and the optical characteristics of the interferometer system.

At least the controller 21 and data processor 32 of the data processing and control apparatus may be implemented by programming computing apparatus, for example a personal computer.

Figure 2 shows a simplified block diagram of such computing apparatus. As shown, the computing apparatus has a processor 25 associated with memory 26 (ROM and/or RAM), a mass storage device 27 such as a hard disk drive, a removable medium drive (RMD) 28 for receiving a removable medium (RM) 29 such as a floppy disk, CDROM, DVD or the like, input and output(l/O) controllers 37 for interfacing with the components of the broadband scanning interferometer system to be controlled by the control apparatus to enable the processor 25 to control operation of these components. The user interface 31 consists, in this example, of a keyboard 31a, a pointing device 31b, a display such as a CRT or LCD display 36a and a printer 36b. The computing apparatus may also include a communications interface (COMMS INT) 199 such as a modem or network card that enables the computing apparatus to communicate with other computing apparatus over a network such as a local area network (LAN), wide area network( WAN), an Intranet or the Internet. In this example, the intensity data receiver 33 is provided as a dedicated frame capture circuit board 230 installed within the computing apparatus.

The processor 25 may be programmed to provide the data processor 32 and controller 2lby any one or more of the following ways: 1) by pre-installing program instructions and any associated data in a non-volatile portion of the memory 26 or on the mass storage device 27; 2) by downloading program instructions and any associated data from a removable medium 29 received within the removable medium drive 28; 3) by downloading program instructions and any associated data as a signal SG supplied from another computing apparatus via the communications interface 199; and 4) by user input via the user interface 31. (

The manner in which the data processor 32 is configured to determine the surface roughness of a thin flhn structure will now be explained.

As a starting point consider a low numerical aperture monochromatic mterferometer (where low numerical aperture means that cos 0 approaches 1 and sinO approaches 0) with the mirror of the measurement arm replaced by a thin film structure The surface height z (that is the path length difference between the measurement arm for that surface pixel and the reference arm) of a surface pixel x, y is represented as a measured phase and is given by: P = + tanIm'1 1) IRe(r)) where ? is the wavelength of the interferometer, and lm(r) and Re(r) are the imaginary and real parts, respectively, of the electric field reflectance r The thin film assembly may be evaluated in known manner. For example, Abelè's matrix approach to the determination of electric field reflectance r may be used as described in, for example, sections 4.8 and 4.9 of "Optical Properties of Thin Solid Films" by OS. Heavens published in 1991 (ISBN 0-486-66924-6). As another possibility, the electric field reflectance r of the thin film structure may be calculated by using the concept of optical admittance as described in, for example, sections 2.3, 2.4, 2.5, 2.6 and 2.7 of "Thin-Film Optical Filters" by [4 Angus Macleod (ISBN 0750306882). Other approaches may also be possible.

From equation 1 above, a perturbation d'P to the measured phase resulting from a perturbation in the surface height z of a surface pixel gives: = E ìzk) + E k k=1 k.1 azk 2) r where dz is the change or perturbation in the substrate surface height and d; is the change or perturbation in the surface height of the k thin film layer and where, for brevity: ( !m(r) x=tan I Re(r) 3) In equation 2), the first two terms represent the phase terms corresponding to the geometrical perturbation and the last term represents the perturbation in the phase change on reflection.

From the above Im(r) tanX= 4) Re(r) Recollecting that the differential of tanx is sec3x and that the differential of u/v is ( Idu Idu' 1 i vt-I--uI-fl--5) dx' dx') v2 then the differential of equation 4) is: Re(r) a(!m(r)) -Im(r) a(Re(r)) 2 az aZk sec x-= _______________________ (3Zk Re2(r) 6) and thus = Re2(r) = Re2(r) Re 2(r) + Im 2(r) R so that 7) . = ! Re(r) a(!m(r)) -tm(r) ä(Re(r))) 8) s9Zk R aZ) For a monochromatic light source of wavelength, the perturbation equation 2) becomes: dP = !Efr + + !er t3(Im(r)) -lm(r) a(Re(r)))) k=1 R 2Zk azk)) 9) Because of the assumption that the interferometer phase I' is exclusively due to surface topography, the general perturbation equation for the apparent surface height or z-perturbation (that is the z-perturbation that would be determined by the data processing and control apparatus) dz1,, for a surface pixel is: dZapp + 1 + 1 (Re(r) a(!m(r)) -Im(r) t3(Re(r)))) dzk = dz + G,Iv,) k 4icv0R äZk)) k1 10) where v0 is the frequency of the wavelength ? of the monochromatic interferorneter system, and Gk is a gain value for the kth layer and represents the amount by which the actual z-perturbation d; for the kth layer has to be multiplied by to obtain the apparent z-perturbation for the kth layer. The gain values Gk correspond, for each thin film, to the rate of change of the apparent surface height with that thin film thickness and the gain values as a set represent the effect on the surface roughness measurements of the actual thin film structure being measured (for example the effect of refractive indices and film thicknesses) and the optical characteristics of the interferometer system, where the optical characteristics are, in this example, the interferometer wavelength and the numerical aperture. In the case being described of a monochromatic interferometer of low numerical aperture, then G=-' k 4az For the simple case of a single non-absorbing thin film, the gain determiner 100 may be configured to evaluate the gain G1(v0) of the single thin film analytically. More generally, the gain determiner 100 will be configured to evaluate the gain values (3k(vO) by numerical differentiation. Thus, for the kth thin film, the gain determiner 100 will evaluate partial ( -.,--11 derivatives of equation 10 as follows: *31m(r) Im(r) -!m(r..2) 1 la) azk and Re(r) -Re(r) dRe(r) lib) where & is a perturbation that represents a very small change in the physical thickness of the kth thin film layer with the value of bz being the same for each thin film layer k. The gain detenniner 100 obtains the electric field reflectances r2 of the k layers by taking advantage of the plane-wave angular spectrum approach discussed in "The distorted helix: Thin film extraction from scanning white light interfrrometry" by Daniel Mansfield published in the Proceedings of SPIE Volume 6186, 618600, 2006 at pages 1 to 11 where the original sand p field reflectances are evaluated using, for example, the standard matrix-based approach as presented in Chapter 2 of the aforementioned text book by Macleod using known or provided values for the substrate and thin film layer refractive indices and the respective nominal thickness of each of the thin film layers, that is the thicknesses specified by the designer of the thin film assembly. Once equations 1 la and lib have been evaluated, then the gain determiner 100 determines the gain values ( in accordance with: 1 + 1 (Re(r) (Im(r)) -Im(r) a(Re(r))) 11 c) 47tv0R t3Zk) The determined gain values are stored by the gain determiner 100.

The manner in which the surface characteristic determiner 101 is configured to determine a surface characteristic will now be described.

When a measurement of surface roughness is made using the apparatus, then the apparent measured surface roughness Scf1, is given by: m Sq = + G(v0) T 12a) where m is the number of layers in the thin film structure and T is the root mean square thickness of the lc thin film layer.

In order to determine the actual substrate roughness Sa and the root mean square thickness Tk of each thin film layer of a thin film assembly having m layers then the surface characteristic determiner 101 is conflgwed to solve mi-i simultaneous equations: Sq,1 1 G G G 2 2 2 2 2 T1 Sq0,,2 1 G21 G....G 2 T = Sq03 i,,,2,-2 "mtl,l m+I,2 "mll.m 2 2 Tm Sq,,p+ i 12b) where G11 to G1, are the gain values stored by the gain determiner 100. Accordingly, the surface characteristic determiner 101 prompts the user to cause the apparatus to make at least m+1 different measurements of apparent surface roughness Sq2,1 to Sq2,1 on m different thin film structure samples that differ from one another only in at least one of the thickness of one or more of the thin film layers and the number of thin film layers. Thus, the material forming the substrate and the materials and the sequence or order of thin film material layers wiibe the same for each sample, although the number of thin film layers and thicknesses differ between samples. Once the m+1 different measurements of apparent surface roughness have been made, then the surface characteristic determiner 101 can solve equation 12b) for the substrate roughness Sq and for each of the root mean square thicknesses T1,T2,T3 Tm because all the gain terms Gj to Gm i m have been determined by the gain determiner 100.

From a purely geometric point of view the front surface perturbation dz is given by: iii 13) so that the corresponding actual top surface roughness is given by: Sq2 = Sq+T k1 14a) The surface characteristic determiner 101 is configured to determine the actual top surface roughness Sq of the surface of the thin film assembly as the sum of the individual mean thin film thicknesses Tk plus the substrate roughness: Sq2= Sq2 + T12 + T....+ Tm2 14b) As described above, the rn-i- i different measurements of apparent surface roughness Sq21 to Sq2,1 are made on m different thin film structure samples. As another possibility, the rn-i-i measurements may be made on a single sample. In this case, the rn-i-i measurements comprise a first measurement of the apparent surface roughness of the uncoated substrate followed by successive measurements of the surface roughness of the thin film layers. This may be achieved by designing the thin film structure so that a measurement region of the substrate and each thin film layer is exposed in the final thin film structure or by making the measurements during manufacture so that the thin film layers are sequentially manufactured and then measured both in terms of apparent surface topography and in terms of mean layer thickness (by, for example, an ellipsometer), one layer at a time. In this case, corresponding surface pixels on successive surfaces (the substrate surface and successive thin film surfaces) should be in registration with one another. If surface pixel registration is maintained throughout the thin film structure, then the surface characteristic determiner 101 may also evaluate the entire set of thickness departures(&(x,y)) from the mean layer thicknesses together with the substrate local departure, Sz1(x,y) in accordance with: 1 0 0 0 1 G, 0 0 Az1(x,y) 1 G31 G32 0.---iz2(x,y) = &2(x,y) 14c) 1 Gm,i,i Gm,i,2 Gmt,m &m(X3Y) 1appm@t*Y) The case of a thin film structure consisting of a single thin film on a substrate will now be considered because it is a type of thin film structure that occurs frequently in practice and represents a case for which the equations are tractable as an analytical expression.

Considering a non-absorbing thin film, then following Abelès and the above-mentioned sections of the afore-mentioned text book by O.S. Heavens: is1 E* = ! 1 T2 1 e Tme (a) 15) E -t2 T2 1 t1 r1e e -I4 o) where r1 and r2 are the electric field reflectance at the film/substrate interface and the air/thin film interface, respectively, E and E are the reflected and incident electric fields, and L1 and t2 are the electric field transmittance at the film/substrate interface and the air/thin film interface, respectively Thus: -iS iS 2 2 -E_ -r1e +r2e -r +r2 +2r1r2cos21 16) ---____________ R -r.r = ___________________ 18 -13 22 e 1-Ti r2e 1 + r1 r2 +2r1 r2cos2&1 Re(r) = r1(1 + r) cos261 r2(1 + r) 17) 1 + rr 2 r1r2cos2S1 r1(1-r)sizi21 18) Im(r) = -______________________ 1 + rr + 2r1r2cos281 Differentiating equations 17 and 18 gives: 19) äRe(r) = 2r1(1 + r)sin2ö1 + 4Re(r)r1 r2sin2ö1 do1 äZ1 1 + rr 2r1r2 cos261 (fr1 äí() = -2r1(1 -r) cos2O1 + 41rn(r) ,1 r2s1n251 ( d5 az1 1 + rr + 2r1r2cos201 d J 20) --15 where the phase: = 2tv0n(v)z1 21a) and d6 = 2irv n(v) dz1 21b) so that, for the low numerical aperture under consideration, the gain expression to be calculated by the gain determiner 100 for the single thin film structure (1(v0) becomes: 1 + 1 R4r) 8(Im(r)) -Im(r) ä(Re(r)) 22) 4jtv0R OZk aZk) and is analytically known.

Figures 3 and 4 show flow charts for illustrating the operation of the apparatus in this case, that is where the apparatus shown in Figure 1 is configured as a low numerical aperture monochsomaticjnterferometer and to carry out phase shifting interferometry where the sample comprises a thin film structure having a single thin film layer and the interfacial surface roughness to be determined is the thin film-ambient atmosphere (usually air) interface surface roughness.

Figure 3 shows a flow chart illustrating one way in which the gain determiner 101 may determine the gain for a thin film structure having a single thin film.

In this example, the gain determiner 100 first determines the operational wavelengtl of the interferometer at SI, the refractive indices of the substrate and the single thin film layer at S2 and the expected or specified mean thin film thickness of the single thin film layer at S3. It will of course be appreciated that the order in which these are determined is not important and that, r for example, the thin film thickness could be determined first. The operational wavelength may be pre-stored by the apparatus or input by the user using the user interface 31 in response to a prompt issued via the user interface 31. The refractive index and mean film thickness will generally be input by the user using the user interface 31 in response to a prompt issued via the user interface 31. The refractive index may be determined by any appropriate conventional manner (for example, ellipsometry or spectrophotometry) or may be supplied by the manufacturer of the sample. The expected or specified mean thin film thickness may be determined by, for example, ellipsometry or spectrophotometry or by using the techniquesdescribed in W02006/005959 or may be supplied by the manufacturer of the sample.

The gain determiner 100 then determines at S4 the surface electric field reflectance r by using the standard matrix relationship as discussed in the aforementioned text book by Macleod.

At S5, the gain determiner 100 calculates the gain G1(v0) of the single thin film either analyticallyor by numerical differentiation as discussed above with respect to equations 10, ha and lib and 22.

The gain G1 is thus dependent upon the refractive index and mean thin film thickness of the single thin film plus the refractive inde,of the substrate and the measurement wavelength of the monochromatic light source, in the example set out above, the mean thin film thickness of the single thin film is supplied by the user. The refractive indices may be supplied by the user or may be determined by the gain determiner 100 from information supplied by the user that identifies the materials forming the substrate and the layer and a data base (which may be local to the apparatus or may be accessed over a network or other communications link) of material refractive indices. The measurement wavelength of the monochromatic lightsource may be pre-stored by the gain determiner 100 or provided by the controller 21.

In the above example, the gain G1 is calculated at the time the user wishes to determine the surface roughness of the thin film structure. As another possibility, the gain determiner 100 or the controller 21 may store a look-up table or similar data set giving the calculated gain G1 for different combinations of substrate refractive index, thin film refractive index and mean thin film thickness, in which case the gain determiner 100 will simply look-up rather than calculate the gain G1 when the user inputs the refractive index and mean thin film thickness As another possibility or additionally, the gain determiner 100 or the controller 21 may store a look-up table or similar data set giving the calculated gain G1 for a number of different thicknesses of each of a number of specific thin film materials. In this case the user may be prompted to input the mean thin film thickness and data identifying the material rather than the refractive index. Of course, if a particular material mayvary in refractive index then the user will be prompted either to input the refractive index or to further identify the material.

If the gain determiner 100 calculates the gain as being approximately zero, then the gain determiner 100 will instruct the controller 21 to advise the user, via the user interface 31, that no gain compensation will be feasible because of the poor signal to noise ratio and, if a different measurement wavelength is available, will advise the user that the measurement wavelength should be changed to change the gain. If this is not possible then the controller 21 will advise the user via the user interface 31 that the interferometric measurement of the thin film sample should be abandoned. -Figure 4 illustrates determination of the interfacial surface roughness for the single thin film thin film structure using the gain (ij determined by the gain determiner 100.

At SlO in Figure 3, the surface characteristic determiner 101 determines a value for the surface roughness Sq2of the substrate. Generally, this will be determined by prompting the user, via the user interface 31, to cause the sample support stage to be moved under the control of the control apparatus to bring a part of the sample free of the thin film into the field of view of the interferometer and then to instruct the control apparatus to carry out a measurement operation in which, for each surface pixel in the field of view, the detector senses the light intensity at each of the number of phase steps and the surface characteristic determiner 101 then determines the phase and thus the surface height for each of those surface pixels in accordance with the usual phase shifting equations as set out in the above-mentioned sections of Optical Shop Testing by Daniel Malacara. The substrate surface roughness Sq2 will then be determined I-' as being the square of the deviation in z of the measured surface pixels.

At Sli, the surface characteristic determiner 101 determines the apparent surface roughness Sq2app for the single thin film of the thin film structure. Generally, this will be determined by prompting the user, via the user interface 31, to cause the sample support stage to be moved under the control of the control apparatus to bring a part of the sample carrying the thin film into the field of view of the interferometer and then to instruct the control apparatus to carry out a measurement operation in which, for each surface pixel in the field of view, the detector senses the light intensity at each of the number of phase steps and the surface characteristic determiner 101 then determines the phase and thus the surface height z, for each of those surface pixels in accordance with the usual phase shifting equations or phase shifting algorithm as set out in above-mentioned sections of Optical Shop Testing by Daniel Malacara. The thin flimsurface roughness Sq21,,, will then be determined as being the square of the deviation in z of the measured surface pixels.

It will of course be appreciated that SlO and Sli could be carried out in reverse order. Also, once the apparent surface roughness measurements have been made for a particular thin film structure sample, then those measurements may be used during the determination of the surface roughness of other thin film structures in produced from the same materials and using the same process conditions which differ only in thin film layer thicknesses.The thin film structure on which the apparent surface roughness measurements are made may be designed to have thin film thicknesses that maximise the signal-to-noise ratio in the gain curve.

At S12, the surface characteristic determiner 101 accesses the gain value G1 determined by the gain determiner 100 and at S13 calculates the actual top surface roughness Sq2. In accordance with equation 14 above, the actual surface roughness is given by Sq2= Sq2+ T12 23) so that, in accordance with equation 12 above, the measured or apparent surface roughness Sq2, is given by:

I

Sq,, = + G(v0) T 24) rearranging equation 24 gives: 2 c'2 app q swbs 25) G21(v,) Thus, at S13 the surface characteristic determiner 101 calculates: 2 Iz 2 -..2 2 -Sq + tGi (vg,) -q -1 -2 26) G1 (v0) and corrects or compensates the surface roughness Sq. As discussed above with respect to equation 14c), if the measurements of apparent surface roughness are made on a single sample and corresponding surface pixels of the substrate and the thin film are aligned, then it may also be possible to determine the surface height z on a surface pixel-by-surface pixel basis.

If the surface characteristic determiner 101 determines that the thin film RMS thickness is much greater (for example approximately three times greater) than the substrate surface roughness: T1 > 27) then the surface characteristic determiner 101 decides that it is reasonable to carry out an approximate calculation to determine Sq as follows: 28) Sq2 Sq,,,/G(v0) (dz tIz,,,/G1(v0)) Figure 5 shows the gain (G1) determined by the gain determiner 100 for a number of sets A to K of thin film structures each of which consists of a single thin film on a substrate having a refractive index of 1.5. The refractive index of the thin films varies from 1.25 for set A to 2.5 for set K and within each set A to K the thin film optical thickness varies from zero to A12 where ? is the wavelength of the monochomatic source of the interferometer system. As a zero film thickness is approached, the gain increases as the refractive index increases. Figure 6 -20 shows the reflectance r as a function of film thickness for the same range of refractive indices (1.25 to 2.5) and a substrate refractive index of 1.5. At a quarter-wavelength thickness, the maximum reflectance is exhibited by the film of highest index.

The above described operation of the gain determiner 100 and the surface characteristic determiner 101 is for a thin film structure having a single thin film layer. The apparatus may be configured to enable the interfacial surface roughness to be determined only for thin film structures having a single layer.

As another possibility, the apparatus may be configured to enable the interfacial surface roughness to be determined for thin film structures having one or more thin film layers. In the latter case, the controller 21 will prompt the user to provide the, the number of thin film layers and the mean film thickness and refractive index for each thin film layer and will provide instructions to the user to enable measurements of the apparent surface roughness of a region of each thin film in the thin film structure in addition to the measurement of the apparent surface roughness of the substrate. As set out above, instead of providing the refractive index, the user may provide information identifying the material so that the refractive index can be obtained from a data base accessible by the apparatus.

For thin film structures have more than one thin film layer, the gain determiner 100 will calculate gain values G1(v0) for the thin film structure in accordance with equation 10 above while the surface characteristic determiner 101 will determine the interfacial surface roughness in accordance with equations 12 and 14 above. Gain values Gk(vO) calculated for thin film structures maybe stored by the gain determiner 100 in association with the details of that thin film structure so that, if the gain determiner 100 determines that the thin film details entered by the user match those of a structure for which the gain values have already been calculated, recalculation is not necessary.

It is clear from equation 12b that there are n-fl unknowns for a thin film assembly with n films and that therefore ni-i measurements will be required. In the case of a two thin-film thin film structure, then from equation 12: Sq,(1) Sq5 + GSq Sq2,,,(2) Sq,+ GT+ GT 29) where G is the gain for the first thin film layer in the absence of the second thin film layer;, G12 is the gain for the first thin film layer when the second thin film layer has been provided on top of the first thin film layer, and G is the gain of the second thin film film layer (the dependence on v has been omitted in equation 29 in the interests of simplicity and because the interferometer in question is a monochromatic one).

Thus, if the data provided by the user indicates that the thin film structure comprises a stack of two thin films, then the gain determiner 100 will carry out calculations in accordance with equation 10 to determine the gains G11, G and G and the controller 21 will advise the user to make measurements of the apparent surface roughness of the uncoated substrate (Sq,3), the first layer only (Sq0(1)) and the surface of the two-film thin film structure (Sq,,,(2)). In order to measure the surface roughnesses Sq,,3 andSq,(1) exposed parts of the substrate surface and first thin film layer need to be brought into the field of view of the interferometer.

These measurements may, but need not necessarily, be made on the same sample. Thus, the measurements may be made on one, two or three samples, provided that the samples are produced by the same process using the same substrate material and the same thin film materials and material layer order. As discussed above, the thickness of a thin film on which a measurement is to be made may be selected so as to optimise the signal to noise ratio in the gain curve.

I

In this case, in accordance with equations 12 and 14, the surface characteristic determiner 101 determines the top surface roughness as: Sq2 T + T Sq = S4(2) + G222_G122Sq2 (1) G -1-G222_G122) Sb$) 30) As discussed above with respect to equation 14c), if the measurements of apparent surface roughness are made on a single sample and corresponding surface pixels of the substrate and the thin films are aligned, then it may also be possible to determine the surface height z on a surface pixel-by-surface pixel basis.

As set out above, once the apparent surface roughness measurements have been made for a particular thin film structure sample, then those measurements may be used during the detennination of the actual top surface roughness of other thin film structures produced from the same materials and using the same process conditions but which may have different thin film layer thicknesses so that it is not necessary to provide measurement access to each thin film layer surface in those other thin film structures.

It will be apparent from equations 12a, 12b, 24 and 29 that the surface characteristic determiner 101 may determine a value for the root mean square thickness of the or one of the thin film layer as an alterative to or in addition to the actual top surface roughness. Such a value may be used, for example, to determine whether the thin film assembly meets manufacturing quality control tolerances.

The explanation above assumes a low numerical aperture interferometer. A typical phase-stepping interfero meter may, however, use a Mirau, Linnik or Michelson interference objective to magnify and image the surface in question. The basic relationship of equation 10 will then be modified to read: Gk(vO) = 1 + 1 Re(r) ä(hn(r)) -Im(r) .3(Re(r)) 1 31) 4Rv0RcosO aZk azk) where: Qis the weighted mean angle of incidence of light from the interferometer on the top surface of the thin film assembly; Gk(vO) is as set out above a function of the electrical field reflectance r which is obtained by taking advantage of the plane-wave angular spectrum approach discussed in "The distorted helix: Thin film extraction from scanning white light interferometry" by Daniel Mansfield published in the Proceedings of SPIE Volume 6186, 618600,2006 at pages ito 11 where the originals and p field reflectances are evaluated using, for example, the standard matrix-based approach as presented in Chapter 2 of the aforementioned text book by Macleod; and in this case the rate of change of apparent surface height with thin film thicknessis given by Gk: Gk= 0cosO 4n aZk Accordingly, where the interferometer system includes such an objective, the electrical field should be integrated over the numerical aperture for that objective to yield a net field reflectance to replace the electrical field reflectance r in equation 34. Otherwise the apparatus functions in the manner described above.

Figure 7 shows a graph of gain G1against thickness in inicrometres (urn) for a silicon thin film on a silicon dioxide substrate where the interferometer system is configured for phase-shifting interferometry and has a measurement wavelength of 450nm (nanometres) and a x50 Mirau interference objective. For a high refractive index film on a low refractive index substrate, the gain maxima occur at half wavelength locations. Figure 8 shows part of the graph of Figure 7 on an enlarged scale. Further results have show that increasing the measurement wavelength causes the gain maxima to decay more gradually with the gain maxima decay being far more gradual for a measurement wavelength of 600am than for a measurement wavelength of 450nm. As mentioned above, the thickness of the layers on which the surface roughness measurements are made may be selected to correspond to a peak in the corresponding gain curve so as to maximise the signal-to-noise ratio.

As described above, the interferometer system is configured as a phase shifting interferometer system. The apparatus shown in Figure 1 may also be configured as a coherence scanning interferometer system. Further details of one example of a coherence scanning interferometer system 2 that may be used in the apparatus I will now be described with the aid of Figure 9 which shows an overall functional block diagram of the apparatus 1, Figure 10 which shows a simplified side elevational view of the mterferometer system 2 and Figure 11 which shows a graph of intensity I against position Z to illustrate an example of the interference fringes (the axial interferogram) produced by coherence scanning interferometry for a sample surface pixel around a coherence peak or interference region.

In the example shown in Figures 9 and 10, the coherence scanning interferometer system 2 is based on a Mirau type interferometer.

As shown in Figure 9, an interferometer I of the broadband scanning interferometer system 2 has a broadband source 4 having first and second components 4' and 4" optically coupled by an optical fibre cable 4b. The first component 4' houses a quartz halogen projector bulb (with associated reflector). The second component comprises a series of optical elements through which light from the optical fibre cable 4b is transmitted. In this example, the series consists of a diffuser, a changeable filter, an aperture stop, a lens, a field stop and a collimating lenhat provides an emergent light beam L. The filter may be a neutral density filter or a band pass filter, designed to restrict the wavelength range of the light emitted by the broadband source 4, such as a Helium-Neon laser line filter designed to pass light having a Helium-Neon laser line wavelength.

The second component 4" of the broadband light source is arranged to supply broadband light L via a beam splitter 12 to an objective lens assembly 13 which includes, in addition to an objective lens 14, the beani splitter S and the reference mirror 6. The beam splitterS splits the light beam provided by the beam splitter 12 into a first reference beam that is directed along the reference path RP and a second sample beam that is directed along the sample path SP from the interferometer I towards the surface 7 of the thin ifim structure mounted on the sample support stage 9. Light returned to the beam splitter 12 is reflected towards the detector by the beam splitter and focussed onto the detector 10 by lens 3 (see Figure 1). q

The objective lens assembly 13, and thus the reference mirror 6, is movable in the Z direction by a Z direction mover 15, in this example a piezoelectric mover, under the control of servo/drive circuitry 15e of the control apparatus 30. The sample support stage 9 is movable in X and Y directions by an X mover 16 and a Y mover 17, respectively, to enable different areas of the sample surface 7 to be brought within the field of view of the detector 10.

As shown in Figure 10, the majority F of the components of the interferometer I of the broadband scanning inierferometer system 2 (apart from first component 4' of the light source and the optical fibre cable 4b)are provided within a housing 2a mounted via a carriage 18 to a2 axis datum column 19. The carriage 18 is coupled via a drive mechanism (not shown) such as a ball screw or lead screw drive mechanism to a coarse Z positioner 20 in the form of a manuallyoperable control or, in this example, a DC motor that enables the carriage 18 and thus the interferometer I to be moved up and down the column 19 in the Z direction to enable the interferometer to be moved to different scanning start positions.

As shown in Figure 10, the sample support stage 9 is provided on a support 102 which houses the X and Y movers 16 and 17. The X and Y movers 16 and 17 comprise, in this example, DC motors coupled to the sample support stage 9 by appropriate conventional drive mechanisms such as rack and pinion or ball screw drive mechanisms (not shown). As shown in Figures 9 and 10, each of the Z, X and Y movers is associated with a corresponding position sensor 15a, 16a and ha while the coarse Z positioner 20 may be associated with a coarse Z positioner position sensor 20a. The dashed lines between the support stage 9 and the X and Y position sensors 16a and 17a in Figure 3 indicate that the position sensors may sense movement of the support stage 9 directly, rather than by signals derived from the corresponding motor.

The controller 21 of the control apparatus 30 controls overall operation of the apparatus and communicateswith the user interface 31 and data processor 32. The control apparatus 30 also includes, in this example, theservo drive circuitry 15e and X, Y and Z loggers 22,23 and 24, each of which receives the output of the corresponding position sensor 16a, 17a and iSa, and a trigger generator 60 for triggering operation of the detector 10 in response to the output of the Z position sensor 15a to capture images at the required intervals. The controller 21 also receives an output from the coarse Z positioner position sensor 20a, if provided. The controller 21 may be programmed in known manner to compensate for any error in the Z position due to the slight arcuate nature of the motion of the objective lens assembly 13.

In operation of the coherence scanning interferometer system, the intensity of the illumination sensed by one sensing element of the 2D sensingarray (that is the portion of the interference pattern provided by light reflected from the corresponding surface region or surface pixel of the sample surface 7 imaged on that sensing element) varies as the path length difference changes with movement of the reference mirror 6, resulting in a series of fringes which have a coherence peak at the position along the scan path corresponding to zero path length difference. Figure 11 shows a graph of light intensity against position Z to illustrate the manner in which the intensity of the light sensed by a sensing element of the 2D sensing array of the detector 10 (and thus the interference fringe region) changes as the relative positions of the reference mirror 6 and sample surface 7 change. The envelope of the intensity distribution is the Fourier transform of the spectral distribution of the broadband source, modified by the spectral transmission of the instrument together with the spectral responsivity of the detector.

The apparatus described above with reference to Figures 9 toll is similar to that described in the applicants international application publication number W003/078925, the whole contents of which are hereby incorporated by reference. Further details of a typical example of the coherence scanning interferometer system 2 and the control apparatus 30 described with reference to Figures 9 and 10 may be found in W003/078925 (our ref 3022799). An example of a commercially available apparatus that may be used is the Talysurf CCI 3000 produced by Taylor Hobson Limited of Leicester, England.

The apparatus being described differs from that disclosed in W0031078925 in the manner set Out above with reference to Figure 1, namely in that it is specifically intended for use in determining surface characteristics of a thin turn structure with a data processor providing, as shown in Figure 1, a gain determiner 100 and a surface characteristic determiner 101.

The manner in which the gain determiner 100 and the surface characteristic determiner 101 operate where the interferometer system is configured as a coherence scanning interferometer system will now be described.

Where the interferometer system is configured as a coherence scanning or broad band interferometer system, then the gain determiner 100 will not calculate the gamin the manner described above because those calculations assume a single wavelength From equation 10, it is apparent that the gain Gk for the k' layer can be expressed by: az Gk= 32) OZk The gain determiner 100 determines the gain values 3k by modelling or simulating the performance of the coherence scanning interferometer system for the thin film structure to be measured using data provided by the user for the thin film to be measured including the substrate complex refractive index and the complex refractive index (n-ik) and the thickness for each thin film of the thin film structure.

lnitially,tbe gain determiner 100 evaluates the net spectral reflectance of the thin film structure in accordance with: G -a 01,1, -33 aZ 2Az where z,, and z,,,, correspond respectively to the simulated interfero meter z-response to having the kth thin film layer incremented and decremented by Az. These apparent surface height changes are derived from the corresponding simulated interferograms through the approach described in W003/078295. Thus from a functional point of view, -34) Z01,1, = Regarding the simulated interferograms, these are derived usingthe relationship presented in --35) I() = R4(e A(v) A(v)e c)) where 1(z1) is the simulated interference intensity at the jth z location along the scan path, the rèflécted electrical field term corresponding to the thin film assembly under the objective, v)e (), is obtained through taking advantage of the plane-wave angular spectrum approach discussed above where the originals and p field reflectances are evaluated using, say, the standard matrix-based approach as presented in Macleod, Thin-Film Optical Filters, the kth layer-thickness perturbation-induced term e 14'is similarly obtained via the plane-wave angular spectrum approach and the instrument optical bandwidth' amplitude term A,(v) is obtained from the measured amplitude of the positive frequency sideband corresponding to the measurement of a spectrally-flat' sample such as BK7 glass.

In summary then, this simulation takes the cs.scntial form of evaluating the net spectral electrical field reflectance, of then generating the corresponding axial-interferogram before finally subjecting this to the procedure described in W003/078295 to extract the apparent surface height, Perturbing the film thickness therefore gives direct access to G.through numerical differentiation.

Figure 12 shows a graph of gain against thin film thickness to illustrate an example of the gain data delemuned by the gain determiner 100 where the interferometer system comprises a coherence scanning interfero meter system with a x50 Mirau objective lens assembly and where the thin film structure comprises a single thin film of silicon on a silicon dioxide substrate. As can be seen by comparing Figure 14 with Figure 8, the main effect of the broadband light source is to integrate out the gain maxima.

Where the inter ferometer system is configured as a coherence scanning interferometer, then the surface characteristic determiner 101 may determine the actual interfacial surface roughness in the manner described above for the case of the monochromatic phase shifting interferometer which requires the user to provide the thin film mean thickness(es).

As another possibility,where the interferometer system is configured as a coherence scanning interferometer, then the surface characteristic determiner 101 may determine the thin film mean thickness(es) using the techniques described in W02006/005959, the whole contents of which are hereby incorporated by reference. Although this can be a relatively slow process, for a multilayer (or single layer) thin film it allows the thin film mean thickness(es) and mean geometric path length difference (which corresponds to z) between the measurement and reference amis to be determined, provided that care is taken to ensure that the substrate and thin films are as flat as possible and are parallel to the reference mirror of the interferometer.

Thus, in this example, the surface characteristic determiner 101 has the functionality of and carries out the operations of the layer structure determiner of W02006/005959 and is operable to request the controller to prompt the user to input the dispersive refractive index n of the substrate and then to cause the apparatus to conduct in either order: 1) a scan of a surface area of the substrate so that a set of intensity data is received by the intensity data receiver 33 for that surface area of the substrate for each scan interval along the scan path; and 2) a scan of a surface area of the surface of the thin film layer assembly so that a set of intensity data is received by the intensity data receiver 33 for thatsurface area of the layer structure for each scan interval along the scan path. -The surface characteristic determiner 101 then carries out a frequency transformation process to provide frequency transformed intensity data for both the substrate and the thin film layer assembly intensity data. In this example, the frequency transformation is a Fourier transformation process and the frequency transformer is arranged to carry out a Fast Fourier Transform (FF1) procedure. The well-known (2) FF1' may be used, although other forms of Fourier transform such as the Winograd Fourier transform (which is not based on powers of 2) could be used.

The surface characteristic determiner 101 then determines, under the control of the controller 21, the ratio between the positive frequency space sidebands of the Fourier transforms and.(I1,,(Z1))55 of the intensity data (if both transforms exceed a threshold, typically set at 5% of their maximum value) for the thin film layer assembly and the substrate, respectively, multiplied by a field reflectance term to provide a normalised HCF(u) ratio. The field reflectance term consists of the difference between the dispersive refractive index of free space (set at 1) and the dispersive refractive index ri(i.) of the substrate divided by the sum of the dispersive refractive index of free space and the dispersive refractive index of the substrate. The surface characteristic determiner 101 can thus calculate the normalised HCF(u) ratio: HCF(v) (1 -n(v)' ..4I(Z,))8, 36) 1 + n(v)) ,I(Z1)) where * indicates the complex conjugate of the bracketed term (the "n base ratio"). Where the substrate is non-absorbing, then the bracketed term will be real and its complex conjugate will be the same. In the interests of generality, however, the reference to the complex conjugate has

been retained throughout this description.

The surface characteristic determiner 101 then sets the calculated normalised HCF ratio equal to an optical admittance ratio term, E(v): HCF(v) a v) e (1 -n(v) .9,(l(z1)58 1 + nb(v)) ,,(1(Z1))8.,.

<p≥ r(v)e 1(V)*4Td(tiZ82)) r(v)e ic(v)e E E(v) 37) where E(v) aE(v) e = 1 -Y'(v)) e 38) 1 +Y'(v)) where * indicates the complex conjugate of the bracketed term, r is the electrical field reflectance from the thin film layer assembly, Y is the optical admittance of the thin film layer assembly(which, as will be explained below, is dependeni upon the following parameters: the number of layers in the thin film layer assembly, the thickness of each layer in the thin film layer assembly and the dispersive refractive index of each layer in the thin film layer assembly), and the exponent is a helical phase term, the derivation of which is explained in W02006/005959.

If the small angle approximation is not appropriate for the NA, then from above the HCF function may be approximated by: r HCF(v) aHCAV)e "h !7 1 -q,(v,O)cosO 1 -q14(v,O)ko6O w(O)dO ___________ 39) 2, 1 1,,(v,O)cosO 1 + q,(v,O)/cosO rve E E(v) where E(v) aE(v)e i(v) = 71 1 -Y;(v,o)co + 1-Y,(V,O)/cosO e 0w(O)dO 40) 2 o ( 1 Y,,'(v,O)cosO 1 + Y,(v,O)/cosO The surface characteristic determiner 101 then requests the user to provide the number of layers, approximate refractive indices and approximate thicknesses of the layers of thin film layer assembly together with an approximate or accurate value for the dispersive refractive index of the substrate and uses these in a thin film optirnisation approach to extract the dispersive refractive index and thickness for each layer of the thin film layer assembly. The fitting procedure of the thin film optimisation approach may be, for example, a conjugate gradient method or a simulated annealing method as discussed in W02006/005959 and as discussed in Numerical Recipes in Fortran: The Art of Scientific Computing, Second Edition' by William H. Press, Saul A. Teukoisky, William T. Vettering and Brian R. Flannery (ISBN 0-521-43064) in section 10.6 at pages 413 to 416 and section 10.9 at pages 436 to 438.

Thus, the local thickness of each layer (that is the thickness corresponding to a selected surface pixel or group of contiguous pixels) and its dispersive index is extracted by setting a ratio, the HCF ratio, related to the ratio between Fourier transforms of the intensity data (in the example described above the ratio of the positive frequency space sidebands of the Fourier Transforms) equal to an optical admittance ratio term that allows the use of thin film fitting procedures to extract the layer characteristics and that makes use of both the amplitude and the phase information available from the measurement operations.

In this case, the surface characteristic determiner 101 determines the HCF ratio or function in the manner discussed above and described in W02006/005959 for a number of different locations within the field of view of the coherence scanning interferometer. The surface characteristic determiner 101 then determines a mean HCF function from the obtained HCE functions.

The mean HCF function is given by: HCF(v) = r(v)e 14Tá,4) 41) and in accordance with W02006/005959, the mean HCF function for a thin film structure may be defined as: HCF(v) = r(v) e = J(v) I e iv) + 42) Considering perturbations of both the thin film layer and the substrate by taking the differential gives: ____ -43) d(HCF(v)) = (-i43rvcos) (r(v)t e "e 4'"'dzS,b$ + (-i4ivcos) (v)I e 1*')e i4WOSO&HcF.fr + (-i) fr(v)J e.ii(v)e 0&M..&d.zk I3Zk so that d(i(v)) = (-i4rvcos) (v) e i(V)e -14J1V03$e&HCF( dz + >.: 1 1 + ________ X dzk) k=1 4mvcosO azk

_ N

<p≥ -i4itvcosOHCF(v) dz,b$ + Gk(v)dzk) 44) using the definition of Gk above.

Thus, the differential of the mean HCF function is gwen by: cHCF(v)) = -i4itvcosHCF(v)( dz+ Gk(v)dzk) 45) In this example, the surface characteristic determiner 101 considers the HCF function for a surface pixel (i,j) as a perturbation from the mean HCF function: HCF(v) = HCF(v) -i4itvcos HCF(v)( tiZ,Jb, + G(v) dzk) 46) k=1 where cosO is 1 for low numerical aperture, and determines, for that surface pixel, the substrate and the thin fun layer thicknesses, dz,,,dz1,dz2,... dzN, by using a standard least-mean squares fitting procedure or other suitable fitting or optimisation procedure. Of course, the Taylor expansion of equation 46 above could be extended beyond the first order which would make the procedure more forgiving in the event of lack of flatness or tilt of the sample being measured.

The surface characteristic determiner 101 repeats this procedure for other surface pixels within the field of view so generating the topography of each interface of the thin film structure, from which the surface roughness Sq can then be derived as discussed above by summing the individual layer topographies and the substrate topography..

As in the case of the monochromatic interferometer, lithe gain determiner 100 calculates any of the gains as being approximately zero, then the gain determiner 100 will instruct the controller 21 to advise the user that no gain compensation will be feasible because of the poor signal to noise ratio and, if a different central measurement wavelength is available for the coherence scanning interferometer, will advise the user that the central measurement wavelength should be changed to change the gain. If this is not possible then the controller 21 will advise the user via the user interface 31 that interferometric measurement of the thin film sample should be abandoned.

The above-described techniques for determining interfacial surface roughness and/or thin film mean thickness T may be used for thin film structures comprising thin films of any reasonably transmissivematerials that enable interference effects to be observed. Generally, the thin films willbe dielectrics or semiconductors. Examples of possible thin film materials include titanium dioxide (Ti02), tantalum oxide (Ta205), silicon dioxide (Si02), Zirconium dioxide (Zr02), silicon.

The present invention may be applied to forms of interferornetry other than those discussed above.

The above description assumes that the spectral range of light source is in the visible. It could, however, extend into or lie within the infra red or ultra violet regions of the spectrum.

The apparatus described above may form part of a thin film manufacturing system. Figure 13 shows a simple block diagram of a thin film manufacturing system 400 comprising thin film designing apparatus 401 that comprises, generally, computing apparatus similar to that shown in Figure 2 programmed: to receive input from a user or operator identifying the features required of the thin film structure such as the required base or substrate, the materials to be used to form the thin film structure, the number of thin film layers and for example the optical characteristics(such as the optical admittance) required of the thin film structure; and to design a thin film structure consistingof layers of the user selected materials having thicknesses that provide the required optical characteristic following, fOr example, the teaching of the afore-mentioned text book "Thin-Film Optical Filters" by H Angus Macleod (ISBN 0750306882).

As another possibility, the user may specify the thicknesses and number of thin film layers in the structure. The thip film designing apparatus 401 provides control instructions for enabling a thin film deposition apparatus 402 to cause the required thin film layer structure to be produced. The thin film deposition apparatus 402 may be configured to form thin film layer(s) 405 on the base or substrate 406 using any known thin film deposition technique(s) such as sputtering, Low Pressure Chemical Vapour Deposition (LPCVD),. Metal Organic Vapour Deposition (MOCVD), Molecular Beam Epitaxy (MBE) and so on. Once the thin film layer(s) have been deposited, the resulting thin film structure 404 is passed to a surface characteristic determining apparatus 403 which is constructed and operates as described above to determine a surface characteristic or characteristics of a produced thin film structure. The surface characteristic determining apparatus 403 may provide control information for at least one of the thin film designing apparatus 401 and the thin film designing apparatus 401 to enable the designing and/or manufacturing process to be modified in accordance with the surface characteristic or characteristics determined by the surface characteristic determining apparatus 403.

Claims

CLAIMS

1. Apparatus for determining mformation relating to a surface characteristic of sample in the form of a thin film structure comprising one or more thin film layers on a substrate, the apparatus comprising: alight source; a light director operable to direct light from the light source along a sample path towards a surface region of a said sample and along a reference path towards a reference surface such that light reflected by a sample surface region and light reflected by the reference surface interfere; a mover operable to effect relative movement along a measurement path between a sample and the reference surface; a sensing device operable to sense light representing the interference fringes produced by a sample surface region; a controller operable to carry out a measurement operation-by causing the sensing device to sense light intensity at a numberof locations along the measurement path to provide, for each location, a corresponding intensity value; and a data processor operable to receive first intensity data comprising a first series of intensity values resulting from a measurement operation on a surface area of a substrate and second intensity data comprising at least a second series of intensity values resulting from a measurement operation on a surface area of a thin film structure, the data processor comprising: a gain determiner operable to determine a gain for the or each thin film of a thin film structure; and a surface characteristic determiner operable to determine a substrate surface characteristic on the basis of the first intensity data, to determine an apparent thin film structure surface characteristic on the basis of the second intensity data, and to modify the apparent thin film structure surface characteristic using the substrate surface characteristic and the gain or gains determined by the gain determiner.

2. Apparatus according to claim 1, wherein the gain represents the rate of change of the apparent surface height with thin film thickness.

-

A. cosO 3. Apparatus according to claim 1, wherein the gain is given byGk = 0 where -4it 8Zk cosO is 1 for low numerical aperture.

4. Apparatus according to claim 1, 2 or 3, wherein the gain determiner is operable to determine the gain for a thin film layer by determining a value for the refractive index and mean thin film thickness of the thin film layer, calculating a surface reflectance and using the surface reflectance and refractive index to calculate the gain.

5. Apparatus according to claim 1, 2 or 3, wherein the light source is a monochromatic light source and the gain determiner is operable to determine the gain for a thin film layer by determining a value for the refractive index and mean thin film thickness of the thin film layer, calculating a surface reflectance and using the surface reflectance and refractive index to calculate the gain. - 6. Apparatus according to claim 1, 2 or 3, wherein the light source is a monochromatic light source and the gain determiner is operable to determine gain in accordance with: dZapp dzWb, + (1 + 4mv0R (Re(r) a(hn(r)) -Jm(r) dz5 + Gk(vO) 12k 7. Apparatus according to claim 1, 2 or 3, wherein the light source is a monochromatic light source and, where the thin film structure has a single thin film, the gain determiner is operable to determine gain in accordance with: 1 + 1 (Re(r) a(Im(r)) -Jm(r) 3(Re(r)) 4ic v0R 8. Apparatus according to claim 1, 2 or 3, wherein the light source is a monochromatic light source and where the thin film structure has a single thin film, the gain determiner is operable to determine gain in accordance with: 22 + + 2r r cos28 R= 1 + rr+2r1 r2cos2&1 2 r(1 + r2)cos2ö1 + 72(1 + r1) Re(r) = ________________________ 1 + rr +2 r1r2cos2&1
2.

rl1-r)S1fl2& fm(T) -___________________ 1 + 2r1r2cos2&1 dRe(r) -2r1(l + r22)sin2&1 + 4Refr)rir2sin2öi( di) i 1 + 1 2 + 2r1r2cos2&1 1 t3fm(r) -2r1(1 -r22)cos261 + 41m(r)r1r2sin2&1 c11 aZ1 1 r r + 2r1r2 cos2ö1 6 where d1 and 61 are respectively the physical and phase thickness of the thin film.

9. Apparatus according to any of claims 1 to 8, wherein the interferometer system has a focussing objective and the gain determiner is operable to determine gain by replacing the surface reflectance with a net field reflectance determined by integrating over the numerical aperture of the objective.

10. Apparatus according to any of claims Ito 9, wherein the controller is operable to cause * the locations on the measurement path to be spaced by a predetermined phase and the surface characteristic determiner is operable to determine a surface height of a surface pixel using a phase shifting algorithm.

11. Apparatus according to any of the preceding claims, wherein the surface characteristic determiner is operable to determine surface roughness.

12. Apparatus according to any of claims 1 to 8, wherein the surface characteristic determiner is operable to determine surface roughness in accordance with: Sq,, = Sq5 + k=1 G(v0) T, and 2 2 2 Sq k-i 13. Apparatus according to any of claims ito 11, wherein, where a thin film structure has a single thin film with a surface roughness much greater than that of a substrate, the surface characteristic determiner is operable to determine surface roughness in accordance with: Sq2 SqIG(v0) 14. Apparatus according to claim 1 or 2, wherein the light source is a monochromatic light source, a thin film structure has two thin films and the surface characteristic determiner is operable to determine surface roughness as: Sq2 = + + Sq5.._{ Sq,,,(2) + G222_G122Sq2 (1) + f G -1-G-G1) S3) 15. Apparatus according to claim 1 or 2, wherein: the light source is a broad band light source; the controller is operable to carry out a measurement operation by causing the sensing device to sense light intensity at a number of locations along the measurement path to provide, for a surface region, a series of intensity values representing -an interferogram in the measurement path direction; and the data processor is operable to determine a surface height of the surface region by determining the position of a coherence peak in the intensity values for that surface region.

16. Apparatus according to claim 15, wherein the gain is given by:.

G = 13Z aZk 17. Apparatus according to claim 15, wherein the gain determiner is operable to determine a gain value for a thin film layer by evaluating a net spectral reflectance for the thin film structure using values; for the layer thicknesses, using the evaluatednet spectral reflectance to generate a corresponding interferogram, determining the position along the measurement path of the coherence peak in that interferogram to obtain an apparent layer thicknesses z,.,, perturbing the layer thickness values by a; to determine the corresponding change az in apparent layer thickness and determining the gain (j on the basis of: G = äZapp äZk 18. Apparatus according to claim 17, wherein the surface characteristic determiner is operable to determine surface roughness.

19. Apparatus according to claim 15 or 16, wherein the surface characteristic determiner is operable to determine surface roughness in accordance with: Sq = + G(v0) T and Sq2 = Sq3+T4 k=1 20. Apparatus according to claim 17, wherein, where a thin film structure has a single thin film with a surface roughness much greater than that of a substrate, the surface characteristic determiner is operable to determine surface roughness in accordance with: Sq2 Sqa,G(vo) 21. Apparatus according to any of claims 17 to 20, wherein the surface characteristic determiner is operable, to enable determination of a surface characteristic, to determine mean thin film thickness(es) by: determining a frequency transform ratio corresponding to a ratio between first intensity data comprising a first series of intensity values resulting from a measurement operation on a thin film layer structure surface area and second intensity data comprising a second series of intensity values resulting from a measurement operation on a surface area of a substrate; and fitting a layer structure model having variable model parameters related to the layer thicknesses and refractive indices of the layers of a thin film structure to the ratio determined by the ratio determiner by adjusting the model parameters, thereby obtaining for the model parameters values representing the layer thicknesses and refractive indices of the layers of the thin film structure.

22. Apparatus according to claim 17, wherein, to enable determination of surface rougbness, the surface characteristic determiner is. operable to determine mean thin film thickness(es) by: determining a frequency transform ratio HCF corresponding to a ratio between first intensity data comprising a first series of intensity values resulting from a measurement operation on a thin film layer structure surface area and second intensity data comprising a second series of intensity values resulting from a measurement operation on a surface area of a substrate; and fitting a layer structure model having variable model parameters related to the layer thicknesses and refractive indices of the layers of a thin film structure to the ratio determined by the ratio determiner by adjusting the model parameters, thereby obtaining for the model parameters values representing the layer thicknesses and refractive indices of the layers of the thin film structure; and wherein the surface characteristic determiner is operable to determine surface roughness by determininga mean of said frequency transform ratios for a number of different surface regions of the thin film structure, determining the differential of the mean frequency transform ratio as follows: d(HCF(v)) = _i47LvcosJfCF(v)( dZ,bs+ Gk(v)dzk) and then carrying out a fitting procedure to determine dz and d;.

23. Apparatus according to aaui 21 or 22, wherein the surface characteristic determiner is operable to determine first frequency transformed data corresponding to the first intensity data and second frequency transformed data corresponding to the second intensity data and to determine the frequency transform ratio by dividing one of the first and second frequency transformed data by the other of the first and second frequency transformed data and by multiplying that ratio by a field reflectance term: HCF(v) (1 -iz,,(v)' + 9,,(f(zk))Sfl ( 1 -n,,(v)1 1 +n,e(v)) 41(Zk)) 1 + nb(v)) where n., (u) is the refractive index at frequency u of the substrate and indicates either the ratio or its complex conjugate.

24. Apparatus according to Claim 21,22 or 23, wherein the ratio determiner is operable to determine: HCF(v) af(v) e tJfC,(V) = ! 7 1 -fl(v,9) cosO + 1 -TIv,O)/cosO O) dO 2 + (v,O) cosO 1 + q(v,0)/cosO r(v)e ic(v)e E(v) 25. Apparatus according to any of claims 21 to 24, wherein the layer structure model is a thin film layer structure optical admittance model.

26. Apparatus according to any of Claims 21 to 24, wherein the layer structure model is a thin film layer structure optical admittance model: E(v) = r (v) e 2liCF = 1 -Y f(v)) e 1CF t. I +Y(v)) where Y indicates the optical admittance, or its complex conjugate.

27. Apparatus according to any of Clamis 21 to 24, wherein the layer structure model is a thin film layer structure optical admittance model: E(v) aE(v)e" = 1 -Y(v,O)cosO + 1 -}(v,O)/cosO e 2 ( 1 + Y,,'(v,O)cosO 1 + Ys*(v,O)/cosO 28. A method of determining information relating to a surface characteristic of sample in the form of a thin film structure comprising one or more thin film layers on a substrate, the method comprising: directing light from a light source along a sample path towards a surface region of a sample and along a reference path towards a reference surface such that light reflected by the sample surface region and light reflected by the reference surface interfere; effecting relative movement along a measurement path between a sample and the reference surface; carrying out a measurement operation by causing a sensing device to sense light at a number of locations along the measurement path to provide, for each location, a corresponding intensity value; and. . processing intensity data comprising a first series of intensity values resulting from a measurement operation on a surface area of a substrate and second intensity data comprising at least a second series of intensity values resulting from a measurement operation on a surface area of a thin film structure by determining an apparent substrate surface characteristicof the substrate on the basis of the first intensity data, determining an apparent thin film structure surface characteristic on the basis of the second intensity data, and modifying the apparent thin film structure surface characteristic using the an apparent substrate surface characteristic and a determined gain or gains.

29. A method according to claim 28, wherein the gain represents the rate of change of the apparent surface height with thin film thickness.

-

?.0cosO 30. A method according to claim 28, wherein the gain is given by: Gk = .-where -4n cosO is 1 for low numerical aperture.

31. A method according to claim 28,29 or 30, which comprises determining the gain for a thin film layer by determining a value for the refractive index and mean thin film thickness of the thin film layer, calculating a surface reflectance and using the surface reflectance and refractive index to calculate the gain.

32. A method according to claim 28,29 or 30, wherein the light source is a monochromatic lightsource and the method comprises determining the gain for a thin film layer by determining a value for the refractive index and mean thin film thickness of thethin film layer, calculating a surface reflectance and using the surface reflectance and refractive index to calculate the gain.

33. A method according to claim 28, 29 or 30, wherein the light source is a monochromatic light source and the method comprises determining the gain in accordance with: dZapp = dz + f (1 + 4R (Re(r) a(Im(r)) -Im(r) ae(r)) = dzbS + I Gk(vO) dzk 34. A method according to claim 28,29 or 30, wherein the light source is a monochromatic light source and, where a thin film structure has a single thin film, the method comprises determining the gain in accordance with: 1 + 1 Re(r) 3(Im(r)) Im(r) a(Re(r)) 4rvR 3z 35. A method according to claim 28,29 or 30, wherein the light source is a monochromatic light source and, where a thin film structure has a single thin film, the method comprises determining the gain in accordance with: r+r+2rrcos2ö R= 12 1 1 + rr+2r1r2cos281 Re(r) = r1(1 r)cos261+r2(1+r) 1 + r12r +2 r1r2cos2ö1 r (1-r)sin2& lm(r)=-122 1 1 + r1 r2 + 2r1r2cos261 8Re(r) -2r(1 + r)sin2&1 + 4Re(r)r1r2sin261 ti1 az1 1 + rr + 2r1r2cos2& alm(r) = -2r1(1 -r)cos261 + 41m(r)r1r2sin2ö1 d1 az1 1 rr + 2r1r2cos2&1 where d1 and 81 are respectively the physical and phase thickness of the thin film.

36. A method according to any of claims 28 to 35, wherein the interferometer system has a focussing objective and the method comprises determining gain by replacing the surface reflectance with a net field reflectance determined by integrating over the numerical aperture of the objective.

37. A method according to any of claims 28 to 36, wherein the locations on the measurement path are spaced by a predetermined phase and a surface height of a surface pixel is determined using a phase shifting algorithm.

38. A method according to any of claims 28 to 37, wherein the surface characteristic is surface roughness.

39. A method according to any of claims 28 to 35, which comprises determining surface roughness in accordance with: Sq = Sq + G(v0) T and m c 2 - 2 -LIq53 Lk k=1 40. A method according to any of claims 28 to 38, wherein where a thin film structure has a single thin film with a surface roughness much greater than thatof its substrate, the method comprises determining surface roughness in accordance with: Sq2 Sq/G(v,) 41. A method according to claim 28 or 29, wherein the light source is a monochromatic light source, the thin film structure has two thin films and the method comprises determining surface roughness in accordance with: Sq2 = T + T+ Sq3 = _{ Sqa,p(2) + G222_G122Sq2 (l)+ (G222 -1-G222_G122)sq2) 42. A method according to claim 28 or 29, which comprises directing light from a broad band light source along the sample and reference paths, carrying out a measurement operation by causing the sensing device to sense light intensity at a number of locations along the measurement path to provide, for the surface region, a series of intensity values representing an interferogram in the measurement path direction, and determining a surface height of the surface region by determining the position of a coherence peak in the intensity values for that surface region.

43. A method according to claim 42, wherein the gain is given by:. G = " k äz

I

44. A method according to claim 42, which comprises determining a gain fora thin film layer by evaluating a net spectral reflectance for the thin film structure using values; for the layer thicknesses, using the evaluated net spectral reflectance to generate a corresponding interferogram, determining the position along the rneasurenient path of the coherence peak in that interferogram to obtain an apparent layer thicknesses Zq,,, perturbing the layer thickness values; by aZk to determine the corresponding change azq, in apparent layer thickness and determining the gain Ci on the basis of:

G

45. A method according to claim 44, wherein the surface characteristic is surface roughness.

46. A method according to claim 42 or 43, wherein the surface characteristic is surface roughness and is.determined in accordance with: m Sq = Sq3 + G(v0) T and Sq2 = Sq5+T kI 47. A method according to claim 44, wherein, where the thin film structure has a single thin film with a surface roughness much greater than that of its substrate, the surface characteristic is surface roughness and is determined in accordance with: Sq2 S4,JG12(v0) 48. A method according to any of claims 44 to 47, which comprises, to enable determination of surface roughness, determining mean thin film thickness(es) by: determining a frequency transform ratio corresponding to a ratio between first intensity data comprising a first series of intensity values resulting from a measurement operation on a thin film layer structure surface area and second intensity data comprising a second series of intensity values resulting from a measurement operation on a surface area of a substrate; and fitting a layer structure model having variable model parameters related to the layer thicknesses and refractive indices of the layers of a thin film structure to theratio determined by the ratio determiner by adjusting the model parameters, thereby obtaining for the model parameters values representing the layer thicknesses and refractive indices of the layers of the thin film structure.

49. A method according to claim 44, which comprises, to enable determination of surface roughness, determining mean thin film thickness(es) by: determining a frequency transform ratio corresponding to a ratio between first intensity data comprising a first series of intensity values resulting from a measurement operation on a thin flimlayer structure surface area and second intensity data comprising a second series of intensity values resulting from a measurement operation on a surface area of a substrate; and fitting a layer structure model having variable model parameters related to the layer thicknesses and refractive indices of the layers of a thin film structure to the ratio determined by the ratio determiner by adjusting the model parameters, thereby obtaining for the model parameters values representing the layer thicknesses and refractive indices of the layers of the thin film structure; and determining surface roughness by determining a mean of said frequency transform ratios for a number of different surface regions of the thin film structure.

50. A method according to claim 44, which comprises, to enable determination of surface roughness, determining mean thin film thickness(es) by: determining a frequency transform ratio HCF correspondingto a ratio between first intensity data comprising a first series of intensity. values resulting from a measurement operation on a thin film layer structure surface area and second intensity data comprising a second series of intensity values resulting from a measurement operation on a surface area of a substrate; and fitting a layer structure model having variable model parameters related to the layer thicknesses and refractive indices of the layers of a thin film structure to the ratio determined by the ratio determiner by adjusting the model parameters, thereby obtaining for the model parameters values representing the layer thicknesses and refractive indices of the layers of the thin film structure; and determining surface roughness by determining a mean of said frequency transform ratios for a number of different surface regions of the thin film structure, as follows: HCF(v)) = -i47rvcosHCF(v)( d+ Gk(v)dzk) and then carrying out a fitting procedure to determine dz and dzk.

51. A method according to Claim 48, 49 or 50, wherein the surface characteristic is determined by determining fist frequency transformed data corresponding to the first intensity data and second frequency transformed data corresponding to the second intensity data, detennining the frequency transform ratio by dividing one of the first and second frequency transformed data by the other of the first and second frequency transformed data, and multiplying that ratio by a field reflectance term: HCF(v) = 1 -n,,(v) .9.41(Zk))S 1 -I + n1,(v)) .j,C4I(Zk)) 1 n(v)) is where n (u) is the refractive index at frequencyu of the substrate and + indicates either the ratio or its complex conjugate. - 52. A method according to Claim 48, 49 or 50, wherein the determined ratio is: HCF(v) a,(v) e PHCAV) = !7 1 -q.(v,O)cos8 + 1 -i(v,O)/cosO 4O)dO 2 1 + i(v,O) cosO 1 + 15(v,O)/cosO r(v) e tq(v) E(v) 53. A method according to any of claims 48 to 52, wherein the layer structure model is a thin film layer structure optical admittance model.

54. A method according to any of Claims 48 to 52, wherein the layer structure model is a thin film layer structure optical admittance model: E(v) = r (v) e I4I&NCF = ( I -Y'(v)) e l+Y(v)) where Y' indicates the optical admittance, or its complex conjugate.

55. A method according to any of Claims 48 to 52, wherein the layer structure model is a thin film layer structure optical admittance model: E(v) a(v)e& = !71 1-1(v,O)cosO 1-Y5(v,O)IcosO e 4HCF" (o)do 2 1 + Y(v,O) cosO 1 + Y(v,O)IcosO 56. Apparatus for determining information relating to a surface characteristic of sample in the form of a thin film structure comprising one or more thin film layers on a substrate, the apparatus comprising: light directing means for directing light from light source means along a sample path towards a surface region of a sample and along a reference path towards a reference surface such that light reflected by a said sample surface region and light reflected by the reference surface interfere; moving means for effecting relative movement along a measurement path between a sample and the reference surface; sensing means for sensing light representing the interference fringes produced by a sample surface region; control means for carrying out a measurement operation by causing the sensing means to sense light intensity at a number of locations along the measurement path to provide, for each location, a corresponding intensity value; and data processing means for receiving tlrst intensity data comprising a first series of intensity values resulting from a measurement operation on a surface area of a substrate and second intensity data comprising at least a second series of intensity values resulting from a measurement operation on a surface area of a thin film structure, the data processing means

A

comprising: gain determining means for determining a gain for the or each thin film of a thin film structure; and surface characteristic means for determining a substrate surface characteristic on the basis of the first intensity data, for determining an apparent thin film structure surface characteristic on the basis of the second intensity data, and for modifying the apparent thin film structure surface characteristic using the substrate surface characteristic and the gain or gains determined by the gain determining means.

57. Data processing apparatus for receiving from an inierferometer first intensity data comprisinga first series of intensity values resulting from a measurement operation on a surface area of a substrate and second intensity data comprising at least a second series of intensity values resulting from a measurement operation on a surface area of a thin film structure, the apparatus comprising: a gain determiner operable to determine a gain for the or each thin film of a thin film structure; and a surface characteristic determiner operable to determine a substrate surface characteristic on the basis of the first intensity data, to determine an apparent thin film structure surface characteristic on the basis of the second intensity data, and to modify the apparent thin film structure surface characteristic using the substrate surface characteristic and the gain or gains determined by the gain determiner.

58. Interferometric data processing apparatus having the features of the data processor set out in any of claims 1 to 27.

59. A signal carrying program instructions to program a processor to carry out a method in accordance with any of claims 28 to 56.

60. A storage medium carrying program instructions for programming processing means to carry out a method in accordance with any of claims 28 to 56.