Comprehensive profile and areal calibration by additively manufactured material measures

The calibration of surface texture measuring instruments is standardized with two distinct types of material measures. ISO 25178-70 categorizes material measures that feature a profile and an areal surface topography. The result is that different types of measuring instruments like profilers on the one hand and areal surface topography measuring instruments on the other hand may require different material measures whose scope of application may be limited to only one of the named instrument types. The reason is that most manufacturing principles allow either a linear or circular extrusion of a geometry limiting the possibilities to manufacture material measures that are suitable for both, profile and areal surface topography measuring instruments. Since a comparability is desired for as many different measuring instruments as possible, we examine to what degree profile and areal material measures of ISO 25178-70 can be adapted and combined to possibly allow a calibration of all types of surface topography measuring instruments. Additive manufacturing with direct laser writing (DLW) is characterized by a high degree of freedom in the design of material measures. An enhancement of structures that can be imaged either in multiple lateral directions or extruded to circular geometries is possible, allowing both, a profile sampling in different directions, just as well as an areal measurement. In the present publication, a modification of the ISO 25178-70 material measures is described including the design process, the manufacturing and the measurement with areal and profile surface topography measuring instruments to practically demonstrate the feasibility of a multifunctional calibration that considers the possible effects of directionality. We show that it is possible to combine different profile and areal geometries by linear and/or areal extrusion of the corresponding profile-based geometry. By aligning multiple material measures onto one sample, it can also be demonstrated that a comprehensive calibration of an optical profiler is enabled with only one measurement.


Introduction and state of the art
The standardization for profile and areal surface texture measurement has emerged independently from each other due to historical reasons.While profile surface texture measurements are established for a long time, areal surface texture measurements emerged starting in the 1980s and 1990s and were standardized starting in 2003 with the foundation of the WG 16 'Areal and profile surface texture' of the ISO TC 213 [1].Also, the field of calibration and the corresponding standardization has emerged individually for each of the two disciplines.The calibration of stylus instruments, which are the typical use case in profile surface texture measurement has been standardized in ISO 12179 [2] and the corresponding material measures that map the characteristics to be calibrated can be found in ISO 5436-1 [3].The geometries and guidelines are customized with regard to stylus instruments, which for a long time was the only common measuring principle for surface texture measurement in the industrial application.Optical surface texture measuring instruments allow both profile and areal measurement.The standardization of areal calibration was embodied in the series of standards ISO 25178 which features the following structure: ISO 25178-70 [4] defines the geometries and characteristics of the material measures applied for the calibration, ISO 25178-600 [5] defines the properties to be addressed in the calibration, referenced as basic metrological characteristics, and ISO 25178-700 [6] defines the procedures for calibration, adjustment and verification.This series of standards addresses several applications since −70 defines both, profile and areal material measures and the basic definitions of −600 and −700 can be applied independently from the measuring principle.Additional standards in the series −601 ff. and −701 ff.define characteristics that are specific to individual measuring principles like stylus instruments, confocal microscopes, or coherence scanning microscopes and the recommended procedures for their calibration.Thus, the ISO standardization framework provides the definitions and routines required for a calibration of both general and specific instrument characteristics.
Due to the measurement time, the application of stylus instruments is in the utmost applications also synonym for a profile measurement, whereas optical measurements can be implemented both using optical profilers of areal surface texture measuring instruments.This means in consequence that the term areal surface texture measurement can also be more or less seen as equivalent for optical measurement.The material measures of ISO 25178-70 are grouped into profile and areal geometries (see figure 1) depending on their application [4].The result is that there is a comprehensive standardization for the calibration of stylus instruments and also for areal surface topography measuring instruments, but no specific guidelines on the use with optical profilers can be found in the standardization.Another consequence is that the categorization into profile and areal material measures also limits the applicability of the individual types of material measures to a certain category of instruments.ISO 25178-70 also contains all types of material measures defined in ISO 5436-1 so that in the long-term it will be withdrawn [7].
The reason for the distinction between profile and areal material measures is that some manufacturing principles can only provide geometries with either a linear or a circular extruded geometry allowing only a limited number of geometries to be manufactured.For profile material measures, there are different methods which have been investigated in previous studies: one typical technology is ultra-precision turning, which is commonly applied for profile material measures featuring a rotational symmetry like chirp material measures [9], depth setting standards (PGC) [10] or roughness artefacts (PCR) [11,12]-and which is a technology that allows the manufacturing of defined profiles with nanometer precision.Material measures that feature a periodic geometry like the periodic triangular (PPT) or the periodic arcuate material measure (PPA) have been generated using etching [13,14].Grinding and similar manufacturing principles are suitable for the generation of PRO Irregular Profile & PCR Irregular Circular Profile material measures [15,16].For the manufacturing of areal material measures, other methods have been investigated, for example lapping [17], milling [18], milling combined with electroforming [19], focused ion beam (FIB) [20] and direct laser writing (DLW) [21], which have all been used for the manufacturing of the type areal irregular surface (AIR) material measure.
Based on the literature it can be concluded that most material measures are manufactured with subtractive manufacturing methods because the still most common tactile sampling requires a high mechanical hardness which is commonly achieved by manufacturing the profile material measures using metallic materials and chipping production.Additive manufacturing methods like the mentioned FIB and DLW are not yet widely established, however, their feasibility to manufacture material measures for optical surface topography measuring instruments has been demonstrated.Besides tactile and optical surface topography measurement, also industrial computed tomography (XCT) is becoming increasingly important in geometrical product specification and due to its nature, it is commonly used to assess the quality of additively manufactured components [22][23][24].In addition to the presented examples illustrating the mapping of defined areal surface texture parameters for AIR-type material measures, additive manufacturing has also been used to manufacture material measures for XCT: Tawfik et al have described an additively manufactured material measure with different internal features to allow a characterization of voids and unfused powder [25,26].Townsend et al demonstrated a round-robin comparison measuring additively manufactured surfaces using XCT [27].Besides classic additive manufacturing, the principles of FIB and DLW have been applied for the manufacturing of material measures.Also, electron beam lithography features a sufficient precision and has been applied for the manufacturing of material measures to determine the instrument transfer function of optical surface topography measuring instruments [28].Additive manufacturing allows many degrees of freedom in the design of material measures for both, profile and areal geometries.Profile and areal material measures have different fields of applications with regard to calibration tasks and as described, due to the rare use of additive technologies, are most often created by different kinds of manufacturing principles.This means as well that the tasks of profile and areal calibration are typically investigated individually.One exception is the chirp material measure, which has been previously modified to allow an application for both, profile and areal evaluation.Profile and areal material measures have different fields of applications with regard to calibration tasks and as described, due to the rare use of additive technologies are most often manufactured by different kinds of manufacturing principles.This means as well that the tasks of profile and areal calibration are typically investigated individually.One exception is the chirp material measure, which has been previously modified to allow an application both for profile and areal evaluation.
Originally, the chirp material measure was designed as a profile material measure [9,29,30], figure 2(a) shows the nominal geometry of a linear chirp material measure and corresponding SEM picture of a manufactured sample by DLW.The manufacturing principles of e-beam lithography [28] an DLW [31] allow the modification to a circular geometry which is shown in figure 2(b).With this circular chirp, a directional evaluation of the transfer behavior of the examined surface texture measuring instrument is possible.Also another areal extension of the profile chirp material measure was proposed: the trapezoidal chirp material measure [32] features continuously changing wavelengths orthogonal to the profile direction.This modified chirp material measure is mapped in four lateral directions.Both named modifications allow a directional profile and an areal evaluation.
To achieve a good comparability and uncertainty estimation for as many types of measuring instruments as possible, a combination of profile and areal material measures featuring identical structures is useful.This was illustrated with the case study of the chirp material measure.One challenge that arises with such a modification is, however, the manufacturing of the designed material measures.Due to the high degrees of design freedom, we examine the modification of profile and areal material measures of ISO 25178-70 to multiple type of measuring instruments and the subsequent manufacturing of the material measures with DLW.Similar to the case study of the chirp material measure, it is possible to image profile material measures as a linear extrusion in different lateral directions to apply a circular extrusion in order to allow an examination of the directionality.The resulting material measures are designed, manufactured and the possibilities for a multifunctional, combined profile and areal calibration with the geometries is examined.The abilities for a comprehensive calibration of optical profilers using a set of material measures on one sample is examined using a combination of circular and linear extrusion of profile-based geometries.The objective is to design a versatile artefact which features as many relevant material measures as possible and thus allows a comprehensive profile calibration with only one profile measurement, while still maintaining the possibilities of an areal calibration of the same characteristics.The examination can illustrate the possibilities that additive manufacturing provides for the combination of profile and areal material measures and also allows a combination of profile and areal calibration, resulting in a better comparability of different measurement systems.

Experimental setup
2.1.Sample design ISO 25178-70 [4] includes 24 types of material measures consisting of 13 profile and 11 areal geometries (see figure 1).It can be observed that there are many geometries that feature similar features, such as periodic groove patterns.Some types of material measures, that are initially applied for the calibration of profile measuring instruments, can be altered to allow a calibration of areal surface texture measuring instruments and the additional axes that is introduced hereby.The objective of the present study is the combination of profile and areal geometries to generate material measures that are applicable for both a profile and an areal calibration.The resulting geometries are suitable for a scanning by optical profilers and areal surface texture measuring instruments.This allows a better comparability of different types of surface texture measuring instruments and can help to reduce the number of different material measures required for a comprehensive calibration of optical surface texture measuring instruments.The result of the design process are 16 combined geometries featuring a size of 100 μm × 200 μm each that are described in the following.
2.1.1.Modification of the periodic material measures types PPS, ARS, CIN, PPR, ACG, PPT, PPA, PAS The design of the modified periodic material measures is shown in figure 3. The following designs were generated: (a) PPS-ARS: A combination of the profile sinusoidal material measure (type PPS) and the areal radial sine wave (type ARS).A radial extrusion of the sinusoidal profile is implemented and split into two segments.The upper segment is linearly extruded resulting in a profile material measure type PPS.When the lower half of the material measure is measured, the type ARS geometry results and the top half can be used for a profile-based sampling of a type PPS material measure.The geometry also allows an extraction of profiles in arbitrary radial directions, enabling the examination of directionality effects of surface topography measuring instruments.The period length is chosen as p = 20 μm and the amplitude as d = 3 μm.
Since there are multiple ways to sample and evaluate the material measures of figure 3, we show some possibilities for the PPS-ARS geometry in figure 4: (i) an areal evaluation of the type ARS material measure is possible when the lower half is measured.The measurands that can be determined are the defined values of Sa and Sq, (ii) an areal measurement of the type PPS geometry, i.e. the upper half of the material measure.The corresponding evaluation parameters are Sa, Sq, PSm.
(iii) a profile extraction of the type PPS material measure and a subsequent determination of Ra, Rq or RSm (iv) the extraction of a profile in a defined direction of the circular geometry.This can be used for the same profile evaluation under the consideration of possible effects of directionality.
(b) Chirp material measure: the design combines both a linear and a circular chirp material measure [28,31] and allows the same four sampling strategies as previously described.This includes the (i) areal sampling of the circular chirp, (ii) areal sampling of the linear chirp, (iii) a profile sampling of the linear chirp, (iv) a directional profile sample of the circular chirp.
(c) Periodic rectangular material measure: the PPR material measure can be combined with the crossed grating ACG by an areal extrusion.The design geometry features a period length of p = 10 μm, a groove width of w = 6 μm, and a depth of d = 3 μm.The evaluation possibilities consist of the determination of all metrological characteristics of ISO 25178-600 defined for the x-and y-axis.These include the amplification coefficients α x and α y , the linearity deviations l x and l y and the x-y-mapping deviations Δ x (x, y), Δ y (x, y) for an evaluation of the ACG material measure [6].The PPR material measure can be evaluated with regard to the evaluation parameters Ra, Rq, RSm, Sa, Sq, PSm.(f) Periodic approximated sine material measure: A similar circular extension of the PAS material measure is designed with p = 20 μm and d = 3 μm and the measurands RSm and Ra, Rq.

Modification of groove material measures types PGR, PGC, PDG, AGC
The next group of modified material measures consists of geometries that feature groove structures.The corresponding designs are summarized in figure 5 with an ongoing subfigure indexing so as not to get confused later: (g), (h) Profile groove circular & Profile groove rectangular: The type PGC and PGR material measure are commonly used for the calibration of the height axis of surface texture measuring instruments and can also be expanded to a circular geometry to allow the acquisition and evaluation of an areal surface topography just as well as profiles in different lateral directions.The groove depths are d = 3 μm.
(i) Profile double groove material measure: The type PDG material measure is also extruded in a circular way so that the areal groove circular (type AGC) material measure results in combination.The grooves feature a distance of l = 50 μm between the grooves and a depth of d = 3 μm.Areal and profile measurements in different directions are possible.

Modification of single-feature material measures types PRI, PRB, APS, APS
The next group of material measures consist of geometries that can be classified as a single feature and are shown in figure 6: (j) Prism: The prism material measure type PRI is extruded in a circular manner, the measurands are the angles between the planes and the parameter Pz which can be evaluated in different lateral directions based on profile-based evaluations.
(k) Razor blade: The type PRB material measure is similarly extruded in a circular direction.The measurand is the measured radius r which can be determined in different lateral directions.
(l), (m) Sphere-Plane and Plane-Sphere material measures: The linear extrusion of the areal geometries type ASP and APS extends the scope of application to profile measurements in different lateral directions.

Modification of roughness material measures types PRO, AIR, ACS
The acquisition of surface texture parameters is commonly calibrated using roughness material measures.The modified versions are summarized in figure 7: (n) Profile roughness material measure: A combination of the areal irregular (AIR) and profile roughness material measure (PRO) is designed by a linear extrusion of the type AIR geometry.This means that both an areal measurement of the AIR and PRO material measure is possible just as well as a profile measurement of the type PRO material measure.In addition to the surface texture parameters Ra, Rq, Sa, Sq the type AIR material measure allows a calibration of the height axis as defined in the standard ISO 25178-600 by the determination of the amplification coefficient α z and the linearity deviation l z of the z-axis based on a defined distribution of height values [33].This information is extracted from the response curve of the height axis.
(o) Areal crossed sinusoidal material measure: The type ACS material measure is also suitable for a calibration of the acquisition of surface texture parameters since ISO 25178-70 defines the surface texture parameters Sa and Sq as measurands.With a linear extrusion a combination with the PPS material measures is possible as shown in figure 7.As design parameters, a wavelength of p = 20 μm and an amplitude of d = 3 μm were selected.
2.1.5.Flatness material measure type AFL (p) Areal flatness material measure: A type AFL material measure with a size of 100 μm × 200 μm is added as well enabling a determination of the instrument noise both for profile and areal measuring instruments (see figure 7).

Arrangement and comprehensive calibration of optical profilers
The 16 previously described material measures are aligned onto one substrate.They are located in a straight line next to each other so that the profile-based sampling of all material measures would be possible by only one measurement.Orthogonal to the profile sampling direction, single lines are printed for the orientation and location during the measurement process.The lateral distance between two material With the layout shown in figure 8, a comprehensive calibration of a profile measuring instrument would be possible by only one profile trace.It is possible to cover all characteristics defined in ISO 12179 [2] for a comprehensive calibration with the aid of only one profile represented by a cross-section through the 16 material measures: the residual profile can be determined with the type AFL (p) material measure, the vertical profile component with the modified groove material measures PGR (g), PGC (h), PDG (i), the   morphological filter effect of the stylus tip which is determined by its radius.Thus, with the given design, a comprehensive calibration of all relevant characteristics of a profile measuring instrument is possible.

Sample manufacturing
The additive manufacturing of multiple material measures onto one substrate is possible with direct laser writing (DLW), as previously investigated in different studies [8,[31][32][33].The advantage of this method is that it does not only allow the large design freedom which is common for additive manufacturing and the possibility to reproducibly manufacture structures with a dimension smaller than the structural resolution of typical industrial optical surface topography measuring instruments [31] but also in comparison to other additive manufacturing principle does not require the application of masks and covers the largest scales of feature size and printing velocity [34].In current research, also the applications of metallic photoresists are examined [35], which in perspective will create more opportunities for an application in the context of material measures.For this application, it is also beneficial that DLW enables the possibility to map multiple material measures onto one substrate [36].Using this advantage, a Universal Calibration Artefact featuring a mapping of all basic metrological characteristics for areal surface topography measuring instruments as defined in ISO 25178-600 can be fabricated in a reasonable time on a single substrate, as proposed in [36].In the present study, a sample which can not only map many different material measures for a calibration of areal surface texture measuring instruments, but also allows a comprehensive calibration of optical profilers is proposed.The samples, designed in section 2.1.6were manufactured by a Photonic Professional GT + device by Nanoscribe GmbH & Co. KG, featuring a 63x objective (NA = 1.4,Zeiss) using Nanoscribe's IP-S photo resist.The slicing distance of the process was chosen with regard to the individual manufactured material measure.For the  material measures that feature dominant flat structures (e.g.PPR, ACG or PGR) a slicing distance of 0.05 μm was chosen to avoid stair-like surfaces caused by the structure optimization.The other material measures that feature dominant non-flat elements were manufactured with a slicing distance of 0.1 μm, since stair-like effects are less prominent when the surface is curved.The hatching distance was chosen as 0.125 μm for all material measures.To reduce the deviations between the nominal and manufactured geometry that are e.g.caused by proximity effects or flatness deviations in the field of view, multiple iterations of the sample were manufactured.After every iteration, an intermediate areal measurement of each surface topography was performed using a confocal microscope to determine the shape deviations between nominal and manufactured geometry and apply a correction to the manufacturing dataset with the objective of reducing the determined deviations.

Measurement
The measurements were done with two different types of surface texture measurements: (i) An areal measurement of each material measure was acquired with the confocal microscope NanoFocus μSurf featuring with a 60x/NA 0.9 objective.The setup features a lateral spacing of 0.226 μm and a nominal vertical resolution of 2 nm.The structural resolution of the setup, representing the wavelength of the smallest structure that can be measured with an amplitude transmission of at least 50%, can be determined in the range of a few micrometers when a series of sine waves is fitted into a measured dataset [32] as suggested by the definition of the small scale fidelity limit [30].
For the measurement of the final geometries, each of the 16 material measures was measured four times, i.e. 64 areal surface topographies were acquired for the characterization of the geometries.The resulting data was used both for a profile-based and areal evaluation to demonstrate the different evaluation possibilities described in section 2.1.
(ii) An optical profiler Precitech CHRocodile using a chromatic confocal working principle featuring nominal parameters of an NA of 0.5, a measurement range of 600 μm, a lateral resolution of 2 μm and an axial resolution of 3 nm.The optical profiler was used to scan 12 repetitive profiles containing a section of all 16 material measures featuring a lateral spacing of 0.5 μm.Subsequently a profile-based evaluation was applied to determine the calibration capabilities of the sample for optical profilers.
Both measuring principles, profile and areal, were selected as measuring principles utilizing geometrical optics to increase the degree of comparability.

Evaluation and results
The final sample after five iterations was first characterized qualitatively using a light microscope and a scanning electron microscope (SEM).The results are summarized in figure 9 where a light microscope image of the 16 material measures (a)-(p) on the substrate and a SEM image showing also the complete layout of the 16 material measures described in section 2.1 and the markers for the positioning in the measurement process.
Subsequently the quantitative topography measurements described in section 2.3 were performed and evaluated with both profile and areal methods.The measurands of the material measures were then determined both for the nominal and measured surface topographies to compare the results and to demonstrate the suitability for both profile and areal calibrations.First, in section 3.1 the evaluations of the areal surface topographies acquired by the confocal microscope are described before in section 3.2, the comprehensive calibration of an optical profiler using the proposed sample is examined.As a general concept, the areal confocal measurement is used for a qualification of the determination of the quality of the structures before they are applied for a measurement by an optical profiler.

Areal measurements using a confocal microscope
The 64 areal datasets were pre-processed by a linear interpolation of the non-measured points and an F-operator consisting of a subtraction of a fitted leastsquares plane.The S-and L-filtering was implemented depending on the type of evaluation.Since from an areal evaluation, four results can be obtained per material measure, the mean value and empirical standard deviation were determined and a 95% confidence interval using the students t-distribution was calculated for all parameters originating from an areal evaluation.Additionally, profiles were extracted from the areal surface topography.The 25 closest profile sections around y = 0.15 mm were used from each of the four areal topographies, resulting in 100 profile evaluations.Based on the 100 evaluated parameter values, a 95% confidence interval can be described assuming a Gaussian distribution with the mean value and standard deviation of all evaluations.The individual evaluations are described in the following categorized like in section 2.1.

Modified periodic material measures PPS, ARS, CIN, PPR, ACG, PPT, PPA, PAS (a)
Type PPS-ARS: The pre-processing of the areal dataset represented by the lower half of the entire material measure did include the mandatory steps of interpolation and F-operator (see above) plus an Sand L-filtering (2.5 μm-250 μm) featuring a linear Gaussian filter (ISO 16610-61).The determined measurands are the amplitude-based surface texture parameters Sa, Sq as defined for the type ARS material measure in ISO 25178-70.Since periodic structures are present, also the autocorrelation length Sal was determined.In the profile evaluation of the upper half of the geometry, a pre-processing of a least-squares straight line removal and a linear Gaussian filter (ISO 16610-21) (ls = 2.5 μm, lc = 250 μm) were used.Then, the profile-based surface texture parameters Ra, Rq of ISO 4287 were determined as measurands.The parameter RSm was calculated using the crossing-the-line algorithm as described in [37].The identical evaluations were applied to the nominal surface topography and the results are compared in table 1.
The parameters feature a good agreement to their nominal values, wheras the areal evaluation features a more notable systematic deviation than the profile evaluation.When the measured surface topography is analyzed as shown in figures 10(a)-(i), a possible reason can be identified by the fact that the areal surface topography is influenced in a more pronounced way by edge effects that are related to the correction in order to compensate the curvature of the field of view introduced by the vignetting of the manufacturing process and the associated structural optimization.Figures 10(a)-(ii) shows the extracted profile at y = 0.15 mm which is in the center of the profile evaluation and (a-iii) an SEM image of the manufactured material measure.
(b) Type CIN: The chirp material measure was evaluated using the topography fidelity, i.e. the determination of the small-scale fidelity limit (SSF) which is defined as shortest wavelength of a structure that is transmitted with a deviation of less than 50% in the given case.In doing so, individual sine waves are fitted and the fitted wavelengths and amplitudes are determined for the description of the instruments transfer behavior.Implausible outliers were removed before the fits were determined.The approach is described in depth elsewhere [30].The small-scale fidelity limit as a 50% transmission threshold is determined in two different lateral directions: for the 25 profiles of each areal surface topography that are located around y = 0.15 mm the left and right half of each profile were analyzed separately to determine the SSF, leading to a total number of 200 evaluations for the four areal topographies.Additionally, the profiles that are orthogonally located in the circular chirp in the lower half were evaluated, leading to two additional evaluations to determine possible effects of directionality and another evaluation based on eight values for the four areal topographies.Figure 11 also illustrates the different evaluation positions and table 2 summarizes the determined results for the SSF.The small-scale fidelity limit shows that in both directions, sinusoidal structures down to about 8 μm have been manufactured and measured.It needs to be point out, that the structures were not optimized with regard to the structural resolution but with regard to the overall compliance to the target geometry.No effect of directionality can be detected.The qualitative agreement between the nominal and the manufactured geometry is well and the circular chirp allows the examination of the transfer behavior in different lateral directions.
(c) Type PPR-ACG, (d) Type PPT, (e) Type PPA, (f) Type PAS: The other periodic material measures were evaluated identical to the type PPS-ARS material measure (a).This means that for the areal surface topography, the parameters Sa, Sq, Sal were determined and for the profile evaluation, Ra, Rq, RSm (see [37]) based on the 25 extracted profiles around y = 0.15 mm for each surface topography.The type ACG material measure was also evaluated with regard to the amplification coefficients and linearity deviations of the x-and y-axis [38].All results are summarized in table A1 (appendix A) as previously described by the provision of a 95% confidence interval.Figure 12 summarizes examples for measures areal surface topographies, the extracted profiles and a SEM image of each type of material measure.
When the results are evaluated, the following observations can be made: the type PPR-ACG material measure (c) features an amplitude which is larger than the desired 3 μm which results in values of the amplitude-based surface texture parameters that are larger than the nominal values.This is caused e.g., by edge Table 1.Results material measure type PPR-ARS (a)-for all results a 95% confidence interval is provided, based on the students t-distribution (areal evaluations, N = 4) or the Gaussian distribution (profile evaluations, N = 100).effects or effects of the correction of the shape deviations and would require further iterations for a convergence.The result of the described correction iterations however is that there is a good qualitative agreement of the structures just as well as a good agreement of the lateral parameters Sal and RSm.This illustrates once more the possibilities for manufacturing that the technology of DLW provides allowing both linear and circular areal extrusions of profile-based geometries.The areal evaluated metrological characteristics of the amplification coefficients and linearity deviations also feature typical values illustrating that a basic areal calibration is enabled with the suggested geometry of the material measure.
The results of the PPT material measure (d) allow similar conclusions: in the extracted profile the edge effects cause by the correction can be observed, also resulting in larger amplitude-based surface texture parameters both in the profile and in the areal evaluation.The lateral characteristics and the qualitative agreement match the expectations well.For the PPA material measure (e), the largest deviations of Sa, Sq, Ra, Rq are present.In the case of this arcuate geometry, one more possible reason for these deviations can be identified in measurement artefacts that are caused due to the steep angles at the peak areas and lead to non-measured points and artefacts in the dataset.The last periodic material measure type PAS (f) confirms the previously described results of a good agreement of the lateral parameters and too large amplitude values   partially caused by edge effects of the correction.The detected deviations are expected to be removable based on more correction iterations of the application of topography measuring instruments that allow a detection of larger surface inclinations.But this was not main target of the present examination.Also, for some type of material measures deviations between e.g. the parameter sets Sa, Sq and their profile equivalents Ra, Rq can be observed.This can be explained by different reasons: first of all, for most of the designed geometries the distribution of height values in the segments for areal and profile evaluation are not identical.As a second reason, the mentioned deviations can be caused by vignetting or over-correction of the structures.A) and examples for the measured surface topographies and SEM images can be found in figure 13.For the single-groove material measures type PGR (g) and PGC (h), the parameters Sv, Rv, Rc were evaluated to determine the groove depth and Svk as an additional measure for the extent of the grooves.The double groove material measure was additionally evaluated with Sal and RSm in order to determine a lateral measure.Rc and RSm were determined as described by the crossing-the-line segmentation [37].Rvmax was determined as an average value for the five sampling lengths.The evaluation of the surface texture parameters just as well as the analysis of the profiles shown in figures 13(g)-(i) and (h-i) show that the depth of the circular groove does not deviate as significantly as the depth of the rectangular groove.The qualitative analysis shows a good agreement to the desired geometries.The results are similar to the previous observations-due to the structure correction, some amplitudes are too large, however a better qualitative agreement was achieved due to the correction and the lateral parameters are in good agreement.It has been demonstrated that both periodic structures and groove structures can be designed as a combination of profile and areal representation.
3.1.3.Modified single-feature material measures-types PRI, PRB, APS, APS As a third category of material measures, material measures with a single, distinct feature are examined.The prism material measure type PRI (j) was examined based on the amplitude-based surface texture parameters Sa, Sq, Ssk, Sku, Ra, Rq, Rku.In contrast, Rsk was not examined as the nominal value of the extracted profiles is 0, making it impossible to determine a percentage deviation.To evaluate the lateral characteristics, the parameter Sal was determined.The parameter results are summarized in table A3 (appendix A), the surface topographies are shown in figure 14.Ssk, Sku, Sal and Rsk feature deviations less than 10%, the remaining parameters are about 20% larger than their nominal values.
The razor blade material measure type PRB (k) was examined using the amplitude-based surface texture parameters and Sp and Rp to determine the peak height of the structures.Rpmax was determined as defined in the crossing the line segmentation [37].Here, the amplitude is slightly smaller than expected which also causes the amplitude-based parameter to feature a negative percentage deviation.The qualitative analysis in figure 14 shows that also the single features can be manufactured with high precision and good agreement to the nominal geometries.
The same also applies for the spherical material measures type ASP (l) and APS (m): a good qualitative agreement can be observed, the parameters shown in table A3 match better for the type APS material measure.This means the extruded sphere material measures match the design quite well even though their measurement is challenging due to the locally large inclinations.The results can be confirmed by the SEM images shown in figure 14.

Modified roughness material measures-types PRO, AIR, ACS
As a modified roughness material measure, the PRO-AIR (n) material measure was designed and evaluated both with the profile and areal surface texture parameters.Additionally, in the areal analysis also the metrological characteristics of the amplification coefficient and the linearity deviation of the z-axis were determined as described in ISO 25178-600.The results in table A4 (appendix A) and figure 15 show a relatively good agreement to the nominal geometry and surface texture parameters with an amplitude that is about 8% larger than the design.The results for the profile and areal analysis are similar.The qualitative agreement is well.
As a second material measure that can be used for a calibration of the areal surface texture parameters is the modified type ARS-PPS material measure.As shown in the qualitative analysis in figure 15, the amplitude is larger than in the design which also reflects in the measures surface texture parameters.The lateral parameters RSm and Sal are however in good agreement.

Flatness material measure type AFL
The results of the flatness material measure type AFL (p) are shown in table A5 (appendix A) and figure 15 (p).Both the profile and areal surface texture parameters (Sa, Sq, Ra, Rq) feature small values of an average of 1 nm, indicating the suitability of the surface for the calibration of instrument noise.
The results of the confocal measurements illustrated that the different type of material measures described in ISO 25178-70 can be manufactured in a modified way-enabling a combined profile and areal calibration with similar results.After the described correction iterations, a good qualitative agreement could be achieved even though some structures could   further be optimized using additional iterations.Direct laser writing was proven to be a suitable manufacturing technology due to the large available freedom of design.In the next step, measurements with an optical profiler were conducted in order to demonstrate the abilities for a comprehensive calibration for this category of measuring instrument.The results are then also compared with the results of the confocal microscope.

Profile measurements and comprehensive calibration
The sample was positioned and in the central area of the profile-based geometries, 12 repetitive profile scans were performed using the optical profiler Precitech CHRocodile.Outliers were removed from the dataset and then all 16 material measures included in one profile were evaluated with the same algorithms as the previously described profile evaluation of the profiles extracted from the surface topographies of the confocal microscope.Figure 16 illustrates one of the measured profiles after the removal of outliers.This profile illustrates the possibility for a comprehensive calibration using only one profile.According to ISO 12179, a comprehensive calibration of a stylus instruments includes the calibration of the residual profile, the vertical and horizontal profile components, the coordinate system, and the entire instrument [2].This is usually achieved by a measurement of a flatness material measure for the determination of the residual profile, the depth setting material measure(s) for the vertical component, different types of periodic material measures for the horizontal component, the prism for the coordinate system and the roughness material measure for the calibration of the entire instrument.Additionally, the razor blade material measure can be used to determine the status and radius of the stylus tip.All named types of material measures are included in the profile shown in figure 16meaning that a comprehensive calibration of a profile measuring instrument is possible based on only one measured profile.The profile can be measured with an optical profiler and all relevant characteristics of the instrument can subsequently be determined.
The measured data of the optical profiler is subsequently compared to the results of the confocal microscope described in section 3.1.The results for the periodic material measures are summarized in figures 17-18 including both a comparison of the measured topographies and evaluated profile surface texture parameters.Depending on the type of material measure and geometry, a varying degree of agreement can be observed.The PPS-ARS material measure (a) features some areas that feature a slope which is too large to acquire for the optical profiler, leading to artefacts in the profile and a detection of approximately half the value of RSm since two profile elements per period are determined instead of one due to the artefacts.For the chirp material measure (b) as well as for the evaluated small scale fidelity limit, similar results of 8-8.5 μm are obtained for both sensors and measurement principles.Also, the periodic rectangular material measure type PPR-ACG (c) features a good agreement with differences of −1.5% to −7.8% between the two sensors.All parameter values and their respective deviations compared to both the nominal values and the measured values of the confocal microscope are summarized in table B1 in the appendix B.
The surface texture parameter evaluated based on the type PPT (d) and type PPA (e) material measure also agree well, even though the optical profiler shows one artefact in the type PPT measurement.The differences in the measurements of the type PAS material measure are more significant due to the artefacts that occur again in the measurement with the optical profiler.The effect is similar than already observed in the measurements of the sinusoidal material measure.The results for the material measures (d)-(f) are summarized in figure 18.
The results of the groove material measures type PGR (g), type PGC (h) and type PDG-AGC (i) are summarized in figure 19 and in table B1 in appendix B. The measurements of the optical profiler exhibit artefacts at the steps located at the edge of all groove structures.This leads as well to deviations of some of the evaluated parameters.Generally, however, a good agreement of the measured topographies can be observed, especially for the type PGR and PGC material measures.
The results for the single-feature material measures are summarized in figure 20 and table B1 in appendix B. The prism material measure (j) shows a very well agreement with parameters only deviating about 3%.In the data of the razor blade material measure type PRB (k), artefacts due to the locally large inclinations can be observed.The height of the sphere material measures type ASP (l) and type ASP (m) agrees well between the two measuring principles even though in the data of the optical profiler artefacts can be observed at the areas with large slopes.
The results of the remaining type of material measures are summarized in figure 21 and table B1 in the appendix B. For the PRO-AIR material measure (n) also a good agreement between the two sensors can be achieved.In the data of the ACS-PPS material measure (o) it is again possible to observe the slope-dependent artefacts in the measurement with the optical profiler which similarly to the other sinusoidal material measures leads to a determination of an RSm value which is about half the nominal value.When measuring the flatness material measure type AFL (p), values of Ra, Rq < 10 nm can be achieved for both measuring principles.The digitization of the data of the optical profiler can be observed in the dataset displayed in figure 21.
To obtain additional information about the transfer behavior of both measuring instruments, the Abbott-Curves of the 16 profile extracts of both instruments shown in figures 17-21 and the corresponding nominal Abbott-Curves are summarized in figure 22.With this comparison the previous results can be confirmed: there are some Abbott-Curves that agree very well between the two measuring instruments and the nominal curve.For some material measures, the deviations possibly caused by vignetting or over-correction in the manufacturing process can be confirmed as a larger amplitude than for the nominal geometry results (e.g.type PPT (d) and type PPA (e)).Some curves also illustrate the different transfer behavior of the two measuring instruments, especially for areas featuring large inclinations.Many Abbott-Curves of the optical profiler show a larger extent of peaks, representing the artefacts causes at the areas featuring large slopes.In direct comparisons of the height distributions, the different transfer behaviors of the two systems can be characterized.As a conclusion, both systems can be calibrated by the same material measures which can expose differences in their transfer behavior.

Conclusions
The technique of direct laser writing allows to break up the strict separation between profile and areal material measures.New material measures were designed by extruding profile-based material measures in a linear and/or circular manner.This led to combined profile and areal material measures that allow a directional evaluation, too.The concept that originates from the circular chirp material measure was transferred to other types of material measures and the general manufacturing feasibility using DLW was demonstrated for the different types of material measures defined in ISO 25178-70.The optimization of the manufacturing data in several iterations led to a good qualitative agreement of the material measures, however some structures will be further optimized with regard to their manufactured amplitudes in future work.
When the areal surface topography measurements using a confocal microscope were evaluated, it could be shown that a good agreement and the possibility for both, a calibration of profile and areal characteristics could be provided.Using reference measurements, it was possible to determine the reference values of the individual material measures that can serve as a base for the further application for calibrations.Deviations in the measured data were also visible at areas featuring a large local inclination.
Using an optical profiler, it could be shown that a single scan is sufficient for a comprehensive calibration.This results from the arrangement of 16 material measures in one row allowing a profile scan of all material measures in only one measurement.The measured data was mostly comparable to the profiles extracted from the areal measurement of the confocal microscope just as well as the measured parameter values.By comparing the height distributions of the Abbott-Curves, differences in the transfer behavior of the two measuring instruments could be exposed and the ability for a combined profile and areal calibration with the same material measures was illustrated.The efficient calibration approach that results for optical profilers is similar to the Universal Calibration artefact that can serve for a comprehensive calibration of areal surface topography measuring instruments [36].With the suggested sample, it is possible to perform a determination of all relevant characteristics of profile measuring instrument based on only one measured profile resulting in a comprehensive calibration of both profile and areal material measures.
Table A1.Results for the periodic material measures Types PPR-ACG, PPT, PPA, PAS-for all results a 95% confidence interval is provided, based on the students t-distribution (areal evaluations, N = 4) or the Gaussian distribution (profile evaluations, N = 100).

Areal evaluation
Profile evaluation

Figure 2 .
Figure 2. (a) Profile chirp material measure, (b) circular chirp material measure, (c) trapezoidal chirp material measure [32].For each type of material measure both the design geometry and an SEM image of the respective DLW fabrications are shown.
(d) Periodic triangular material measure: The type PPT material measure is circularly extruded, enabling an areal measurement of either the upper or lower half, or an extraction of profiles in different lateral directions.The profile-based evaluation is based on the parameters Ra, Rq, RSm, Sa, Sq, PSm.The design parameters are set as period length p = 20 μm and amplitude d = 3 μm.(e) Periodic arcuate material measure: The type PPA material measure was varied in an identical way like the PPS, PPR and PPT material measure also featuring p = 20 μm and d = 3 μm.
horizontal profile components with the modified periodic material measures type PPS (a) PPR (c), PPT (d), PPA (e), PAS (f), the profile coordinate system with the modified single feature material measure prism PRI (j), sphere-plane ASP (l) and plane-sphere APS (m) are suitable and the overall instrument can be calibrated by the evaluation of the modified roughness material measures type PRO (n), PPS (o) (a), PPR (c), PPT (d), PPA (e), PAS (f) by a determination of Ra, Rq.In addition, the chirp material measure CIN (b) allows a determination of the transfer behavior and the razor blade PRB (k) can classify the morphological filtering caused by the finite size of a lightspot-similar to the

Figure 8 .
Figure 8. Layout of the substrate containing the 16 designed material measures (a)-(p).

Figure 9 .
Figure 9.Light microscope image of the manufactured substrate (top) and SEM image (bottom).

Figure 11 .
Figure 11.Chirp material measure-measurement data.Respective cross-sections at y = 0.15 mm are shown in the upper right corner For each material measure an areal surface topography from confocal measurement, evaluation positions (i), a cross-section at y = 0.15 mm (ii) and an SEM image (iii) are shown.The evaluation positions are indicated.

Figure 12 .
Figure 12.Periodic material measures-measurement data.(c) type PPR-ACG, (d) type PPT, (e) type PPA, (f) type PAS.Respective cross-sections at y = 0.15 mm are shown in the bottom.For each material measure an areal surface topography from confocal measurement (i), a cross-section at y = 0.15 mm (ii) and an SEM image (iii) are shown.

Figure 13 .
Figure 13.Groove material measures-measurement data.(g) type PGR, (h) type PGC, (i) type PDG-AGC.Respective cross-sections at y = 0.15 mm are shown in the bottom.For each material measure an areal surface topography from confocal measurement (i), a cross-section at y = 0.15 mm (ii) and an SEM image (iii) are shown.

Figure 14 .
Figure 14.Single-feature material measures-measurement data.(j) type PRI, (k) type PRB, (l) type ASP, (m) type APS.Respective cross-sections at y = 0.15 mm are shown in the bottom.For each material measure an areal surface topography from confocal measurement (i), a cross-section at y = 0.15 mm (ii) and an SEM image (iii) are shown.

Figure 15 .
Figure 15.Roughness material measures and flatness material measure-measurement data.(n) type PRO-AIR, (o) type ACS-PPS, (p) type AFL.Respective cross-sections at y = 0.15 mm are shown in the bottom.For each material measure an areal surface topography from confocal measurement (i), a cross-section at y = 0.15 mm (ii) and an SEM image (iii) are shown.

Figure 16 .
Figure 16.Measured profile of the material measure with the optical profiler.

Figure 17 .
Figure 17.Comparison of periodic material measures-measurement data.(a) type PPS-ARS, (b) type CIN, (c) type PPR-ACG.Profile topography data of confocal microscope and optical profiler and comparison of the evaluated surface texture parameters.A 95% confidence interval is shown with the error bars for the confocal microscope (N = 100 except of SSF with N = 200) based on the Gaussian PDF, for the optical profiler (N = 12 except of SSF with N = 24) based on the student's t-distribution.

Figure 18 .
Figure 18.Comparison of periodic material measures-measurement data.(d) type PPT, (e) type PPA, (f) type PAS.Profile topography data of confocal microscope and optical profiler and comparison of the evaluated surface texture parameters.A 95% confidence interval is shown with the error bars for the confocal microscope (N = 100) based on the Gaussian PDF, for the optical profiler (N = 12) based on the student's t-distribution.

Figure 19 .
Figure 19.Comparison of groove material measures-measurement data.(g) type PGR, (h) type PGC, (i) type PDG-AGC.Profile topography data of confocal microscope and optical profiler and comparison of the evaluated surface texture parameters.A 95% confidence interval is shown with the error bars for the confocal microscope (N = 100) based on the Gaussian PDF, for the optical profiler (N = 12) based on the student's t-distribution.

Figure 20 .
Figure 20.Comparison of single-feature material measures-measurement data.(j) type PRI, (k) type PRB, (l) type ASP, (m) type APS.Profile topography data of confocal microscope and optical profiler and comparison of the evaluated surface texture parameters.A 95% confidence interval is shown with the error bars for the confocal microscope (N = 100) based on the Gaussian PDF, for the optical profiler (N = 12) based on the student's t-distribution.

Figure 21 .
Figure 21.Comparison of roughness and flatness material measures-measurement data.(n) type PRO-AIR, (o) type ACS-PPS, (p) type AFL.Profile topography data of confocal microscope and optical profiler and comparison of the evaluated surface texture parameters.A 95% confidence interval is shown with the error bars for the confocal microscope (N = 100) based on the Gaussian PDF, for the optical profiler (N = 12) based on the student's t-distribution.

Figure 22 .
Figure 22.Comparison of the Abbott-Curves for the profile topography data of confocal microscope and optical profiler.