Aus Wissenschaft und Forschung 5 Tribologie + Schmierungstechnik · 70. Jahrgang · 1/ 2023 DOI 10.24053/ TuS-2023-0002 Application of batch manufactured flexible micro-grinding tools on copper and oxidized copper surfaces Lukas Steinhoff, Folke Dencker, Marc Christopher Wurz* Eingereicht: 18.8.2022 Nach Begutachtung angenommen: 7.2.2023 Dieser Beitrag wurde im Rahmen der 63. Tribologie-Fachtagung 2022 der Gesellschaft für Tribologie (GfT) eingereicht. Kupfer ist im Vergleich zu Aluminium aufgrund seiner Duktilität ein eher schwierig zu zerspanendes Material. Deshalb wird in dieser Studie der Ansatz verfolgt, die Oberfläche zu oxidieren, um die Zerspanbarkeit beim Schleifprozess und somit die erreichte Oberflächenqualität zu verbessern. Dazu werden batchgefertigte, flexible Mikroschleifwerkzeuge für das Schleifen von Kupfer und oxidierten Kupferoberflächen verwendet, um insbesondere Flächen von Mikrostrukturen oder lokal kleine Areale von Flächen bearbeiten zu können. Außerdem wird ein Vergleich der Schleifergebnisse bei der Nutzung der Abrasive Siliziumkarbid (SiC) und kubischem Bornitrid (cBN) als Schleifmittel gezeigt. Die Werkzeuge bestehen aus einer Schleifschicht auf Polyimidbasis und den Abrasivpartikeln auf Silizium als Trägersubstrat und werden durch Fotolithographie und reaktives Ionentiefenätzen hergestellt. Die Oxidation der Kupferoberflächen findet durch einen elektrochemischen Prozess statt und anschließend werden diese mit den hergestellten Schleifwerkzeugen bearbeitet. Die Oberflächenqualität wird durch optische Messungen mittels konfokaler Lasermikroskopie hinsichtlich der Rauheit bewertet. Sowohl auf dem reinen Kupfer als auch auf dem oxidierten Kupfer werden durch den Kurzfassung Einsatz von SiC-Schleifwerkzeugen geringere Rauheitswerte erzielt. Dies spiegelt sich bei reinem Kupfer in einer Reduktion des arithmetischen Rauheitsmittelwertes R a auf 0,04 µm wider. Die unbearbeitete Referenzfläche weist einen R a -Wert von etwa 0,24 µm auf. Darüber hinaus zeigen die bearbeiteten oxidierten Oberflächen eine Reduktion der mittleren Rautiefe R z von 7,60 µm auf etwa 1,10 µm, was im Vergleich zu den bearbeiteten nicht oxidierten Kupferoberflächen (2,32 µm) etwa eine Optimierung um Faktor 2 darstellt. Die Bearbeitung von Kupfer mit cBN-Mikroschleifwerkzeugen zeigt ebenfalls reduzierte Rauheitswerte, die jedoch im Vergleich zu den SiC-Werkzeugen bei bearbeiteten Kupferoberflächen um 50 % höher und bei bearbeiteten oxidierten Kupferoberflächen ähnlich sind. Während sich die Oxidation der Kupferoberfläche positiv auf die Oberflächenqualität auswirkt, ist kein Einfluss auf den Werkzeugverschleiß zu beobachten. Schlüsselwörter Anodische Oxidation, Mikroschleifen, Feinmechanik, Hochpräzisionsbearbeitung, Kupferbearbeitung As copper is a rather difficult material to machine due to its ductility compared to aluminium, this study presents the approach of oxidizing the surface to improve the results of the grinding process. Therefore, batch manufactured flexible micro-grinding tools are used for grinding of copper and oxidized copper surfaces to machine microstructure or local areas of functional Abstract surfaces. Besides, we show a comparison of the performance of an abrasive layer made of silicon carbide (SiC) and cubic boron nitride (cBN). The tools are made of a polyimide-based abrasive layer and silicon as substrate and are fabricated by photolithography and deep reactive ion etching. The oxidation of copper surfaces is done by electrochemical processes and are * Lukas Steinhoff, M. Sc. 1 Orcid-ID: https: / / orcid.org/ 0000-0003-3411-5802 Folke Dencker, M. Sc. 1 Orcid-ID: https: / / orcid.org/ 0000-0002-9917-0565 Prof. Dr.-Ing. Marc Christopher Wurz 1,2 Orcid-ID: https: / / orcid.org/ 0000-0002-2066-8142 1 Institute of Micro Production Technology Leibniz University Hannover Germany 2 DLR Institute of Quantum Technologies Ulm University Germany crease of the machinability due to higher brittleness and thus enables the achievement of higher surface qualities. Anodic oxidation of copper allows the creation of nanoneedle structures on the surface. These are more brittle und mechanically unstable [7]. 2 Experimental procedures In the following, the production of the tools with SiC and cBN as abrasive grains and the anodic oxidation of copper is described first. Then, the description of the grinding process and the analysis of the machined copper follows. 2.1 Production process The grinding tools are composed of two parts, the tool head and the tool shaft. The shafts are milled out of aluminium. The tool heads are batch produced using microsystem technologies. The process is shown in Figure 1. The polyimide-abrasive-suspension (PAS) consists of 25 wt% SiC or cBN particles with a size of 3-6 µm and 75 wt% of a polyimide precursor, which are mixed evenly. The weight distribution was determined to be optimal in previous tests. The mixing is followed by spin coating, a lithographic patterning to create round structures with a diameter of 1 mm and deep reactive ion etching (DRIE) to reach a structural height of around 250 µm of each tool head. After dicing, tool heads are joined to tool Aus Wissenschaft und Forschung 6 Tribologie + Schmierungstechnik · 70. Jahrgang · 1/ 2023 DOI 10.24053/ TuS-2023-0002 1 Introduction Due to advancing technological progress, high-quality surfaces are becoming more and more important. In particular, the surface profile can strongly influence the efficiency of a functional surface. For example, undesirable roughness in optical components can lead to scattering [1]. Moreover, technological advances are also accompanied by miniaturization and the size of the functional surfaces shrinks. The machining of such small surfaces and microsystems in general, falls into the field of micromachining, which includes grinding as a mechanical machining method [2]. Therefore, suitable tools are needed to finish the manufactured components in order to increase the surface quality, especially for commonly used materials such as copper. It is used in optics as mirror material and electronics thanks to its excellent properties, such as high thermal conductivity or high electrical conductivity, but does not show a good machinability [3]. Moreover, a flexible binding matrix is favourable, because it yields while machining and so the component is kept true to contour [4]. Grinding wheels with polyimide as flexible binding matrix have already been demonstrated in macroscopic size to successfully machine steel [5]. Furthermore, micro-grinding tools on the same basis with silicon carbide (SiC) as abrasive have already been successfully used for machining to a mean roughness of R a = 0,04 µm and a mean roughness depth of R z = 2,4 µm [6]. In this study, cubic boron nitride (cBN) as abrasive is compared with SiC for optimizing the machining of copper. In addition, an electrochemical oxidation of the copper surface shall lead to an indirectly machined with grinding tools. The surface quality is evaluated concerning the surface roughness by optical measurements with confocal microscopy. Lower roughness values are achieved on both, the pure copper and the oxidized copper by using SiC grinding tools. On pure copper this is reflected in a reduction of the arithmetical mean roughness value R a to 0.04 µm. The unprocessed reference surface shows an R a of 0.24 µm. In addition, the machined oxidized surfaces show a reduction of the mean roughness depth R z from 7,60 µm to 1.10 µm, which is an optimization of factor 2 compared to the machined non-oxidized copper surfaces (2.32 µm). The machining of copper with cBN micro-grinding tools also shows improved roughness values, but in comparison to the SiC tools these are 50 % higher for machined copper surfaces and similar for machined oxidized copper surfaces. While the oxidation of the copper surface has a positive effect on the surface quality, no effect on tool wear can be observed Keywords Anodic oxidation, micro-grinding, precision engineering, high precision machining, copper machining Figure 1: Production process of micro-grinding tool heads. It is separated into mixing of the suspension (a), spin coating of the PAS onto silicon substrate (b), structuring by lithography and DRIE (c) and separation by dicing (d). shafts using 2K epoxy adhesive (Plus Endfest 300, UHU). More detailed information about the process and the effects of the process parameters on the tool are described in a previous publication [6]. For the electrochemical modification of copper, shown in Figure 2, the workpieces are machined to a size of 25x25x6 mm3. Furthermore, the surface is roughened manually with FEPA P 1200 sandpaper to create a reference surface. Before electrochemical modification, the copper is cleaned with solvents and the native oxide layer is removed using 37 % hydrochloric acid. Subsequently, anodic oxidation takes place in a 2 mol/ l sodium hydroxide solution in a three-electrode configuration. The copper is the working electrode (WE). A platinized titanium grid is used as counter electrode (CE) and an Ag/ AgCl electrode as reference electrode (RE). The potential is controlled by a potentiostat, which keeps the potential at - 200 mV. The process time is 10 min, resulting in an average layer thickness of 5 µm. After the anodic oxidation, the copper is removed from the electrolyte, cleaned with distilled water and dried with nitrogen. After successful anodic oxidation, a light blue coating on the copper workpiece can be observed. The topography can be analysed by scanning electron microscopy (SEM). The layer thickness is measured by confocal microscopy. In addition, the oxidized layer is examined for its polycrystallinity and layer composition by means of X-ray diffraction (XRD) measurement. 2.2 Grinding process The grinding tools with embedded SiC or cBN particles are first tested on pure copper surfaces. For this purpose, the copper is milled flat and roughened by FEPA P 1200 sandpaper in order to create the reference surface. Subsequently, a circular structure with a diameter of 4 mm and an area of 12.57 mm 2 is ground with a tool path length of 42.14 mm. For this purpose, a cutting speed of 62 m/ min, a feed rate of 100 mm/ min and an infeed of 5 µm are used. A grinding strategy is pursued in which the surface is ground only once, as well as one in which the surface is ground four or seven times. After each grinding step, the tool wear is measured so that the infeed is kept as equal as possible. The machine tool used is a 5-axis CNC machine (C5, Datron). The surface roughness is analysed using confocal microscopy. In addition, the tool wear is determined for a tool path of 880 mm, where the tool wear that has occurred is measured every 40 mm. Tactile measurement methods are used to determine the layer thickness of the removed material to determine the g ratio (Removed volume workpiece / lost volume grinding tool). The grinding strategies described above are also applied to the oxidized copper surfaces. Subsequently, the same analysis strategy is applied to compare the reachable surface roughness on machined copper and machined oxidized copper. 3 Results and discussion This part is divided into the two main points of the paper. First, the two types of tools are compared in the grinding process on pure copper. Then, the difference in the achievable surface finish on machined copper and machined oxidized copper is analysed. 3.1 Comparison of SiC and cBN micro-grinding tools To compare the two types of tools, the surface quality produced is compared on the one hand and the tool wear and the g ratio on the other. Figure 3A shows the surface roughness measured on the machined copper. For R a and R z , a significant improvement was achieved with both tool types compared to the roughness on the reference surface (0 grinding steps). Although the same reference surface is used, it can be seen that the use of the SiC tools lead to a lower surface roughness than the use of the cBN tools. For the latter, a total reduction of the R a value from 0.24 µm to 0.06 ± 0.01 µm and of the R z value from 7.60 µm to 3.31 ± 1.01 µm can be observed. Under the same conditions, the SiC tools achieve an R a value of 0.04 ± 0.02 µm and an R z value of 2.32 ± 0.59 µm. However, after only one grinding step, an improvement in surface finish (R a = 0.08 ± 0.02 µm, R z = 3.49 ± 0.83 µm) can be observed when machining with SiC tools. Increasing the number of grinding steps to four then leads to a further improvement (R a = 0.05 ± 0.02 µm, R z = 2.70 ± 0.53 µm), as shown in Figure 3A. This observation can also be made for cBN tools. Here, the first reduction of the surface roughness is already achieved after one grinding step (R a = 0.11 ± 0.03 µm, R z = 5.22 ± 0.69 µm) and an am- Aus Wissenschaft und Forschung 7 Tribologie + Schmierungstechnik · 70. Jahrgang · 1/ 2023 DOI 10.24053/ TuS-2023-0002 Figure 2: Production process of oxidized copper workpieces. It is separated in preparation (a), anodic oxidation (b) and cleaning (c). tools and 16.5 µm for cBN tools. A difference between the two tool types cannot be seen. This suggests that the wear behaviour is predominantly controlled by the properties of the bonding matrix, as this is the same for both tool types. The g ratio is also in favour of the SiC tools. It results from the ratio of the removed workpiece volume to the tool volume and results in a value of 52 for SiC and 31 for cBN tools. Overall, the results indicate that the SiC tools are more suitable for machining pure copper workpieces, as the achieved roughness values R a and R z are both lower with nearly the same tool wear. 3.2 Comparison of machinability of copper and oxidized copper surfaces After shown the comparison of SiC and cBN microgrinding tools on pure copper surfaces, it can be seen, Aus Wissenschaft und Forschung 8 Tribologie + Schmierungstechnik · 70. Jahrgang · 1/ 2023 DOI 10.24053/ TuS-2023-0002 plification of this effect occurred after four grinding steps (R a = 0.10 ± 0.04 µm, R z = 3.77 ± 0.82 µm). In SEM images, a more heterogeneous size distribution for SiC can be observed. The comparison is shown in Figure 4, which includes cBN grains in Figure 4A and SiC grains in Figure 4B. Some of the SiC grains have an edge length of up to 10 µm in one direction. As a result, the grains showed larger cutting edges, which could lead to better cutting performance and therefore to a better surface finish. In terms of tool wear, both tools show similar behaviour, as can be seen in Figure 3B. In the first 120 mm, a tool wear of about 0.09 µm/ mm and a total tool wear of 11.13 µm for SiC tools and 11.23 µm for cBN tools can be observed. Subsequently, the tool wear decreases to 0.008 µm/ mm and the total tool wear increases less rapidly. After 880 mm, it increases to 16.7 µm for SiC Figure 3: Comparison of results after using SiC and cBN grinding tools. In A, the average (n = 3) surface roughness values achieved with the respective SiC and cBN tools are compared. B shows the total tool wear curves over a total tool path length of 880 mm for both tool types. Figure 4: Comparison of the size distribution of integrated cBN (A) and SiC (B) particles. Despite the specified size distribution of 3-6 µm for both particle types, a more heterogeneous size distribution is noticeable for the SiC grains than for the cBN ones. A B that the reachable surface roughness under those conditions is limited. One approach to improving the surface finish is to modify the copper surface. Oxidized copper is less ductile and therefore easier to machine. In this case, the oxidation was carried out electrochemically as previously described. The resulting surface, as shown in Figure 5A, is very rough with small needle structures. Compared to the otherwise used reference surface, this modification initially leads to an increase in surface roughness, as can be seen in Figure 5B. An R a value of 0.64 µm and an R z value of 9.08 µm are obtained. XRD measurements show a polycrystallinity of the layers as well as a proportion of copper hydroxide through peak detection, which is in line with the literature [8, 9]. Due to the multi-step machining the roughness decreases in a similar way as before on the pure copper. However, the two types of tools show similar efficiency on oxidized copper. Furthermore, the R z value is significantly lower in both cases after seven grinding steps than on pure copper, while the R a values are nearly the same. Previously, R z was 2.32 ± 0.59 µm for SiC tools, whereas the value is now less than half that, namely R z = 1.10 ± 0.35 µm. This could be a consequence of the electrochemical process. Similar to electropolishing, a higher current density could prevail at the roughness peaks, favouring a stronger reaction. In addition, it has been shown in the literature that the generation of the nanoneedle structures is controlled mainly by current density [9], which suggests also a stronger reaction at the roughness peaks. 4 Conclusion In this work, the suitability of different abrasive types for machining copper was tested. SiC showed the better performance than cBN based tools, resulting in average R a values of 0.04 ± 0.02 µm and R z values of 2.32 ± 0.59 µm after seven grinding steps. The reason for this could be the different cutting edge geometry of the abrasive grains. The g ratio (52) was also higher for the SiC tools due to higher material removal. To further improve the surface finish, a modification of copper by anodic oxidation was tested, which resulted in a needleshaped surface structuring. Both tool types showed similar results here. In particular, a successful improvement was observed when considering the R z value. Compared to the machined copper, the R z value for SiC tools was more than halved, namely 1.10 ± 0.35 µm. The results indicate that a precise oxidation of the copper surface can lead to better grinding results. Therefore, it can be advantageous to precede a grinding process with a surface modification of the copper workpiece by anodic oxidation. For a good integration into the existing CNC process, it could be advantageous to develop an electrochemical chamber for the machine tool. In addition, it was confirmed that the developed micro-grinding tools can not only be used on pure copper surfaces, but are also functional on oxidised copper surfaces. This means that the tools, in combination with the oxidation process, can produce a high surface finish through a simple grinding process on copper. Acknowledgement The project “Batch processed flexible micro-grinding tools for end machining of metallic surfaces” (WU 558/ 26-1) is financial supported by German Research Foundation (DFG). References [1] Harvey J E, Lewotsky K L, Kotha A 1995 Effects of surface scatter on the optical performance of x-ray synchroton beam-line mirrors Appl. Opt. 34 3024-3032 [2] Brinksmeier E and Preuss W 2012 Micro-machining Phil. Trans. R. Soc. A 370 3973-3992 Aus Wissenschaft und Forschung 9 Tribologie + Schmierungstechnik · 70. Jahrgang · 1/ 2023 DOI 10.24053/ TuS-2023-0002 Figure 5: Results of grinding oxidized copper. A shows an SEM picture of the oxidized copper surface. B illustrates the surface roughness progress after grinding with SiC and cBN tools. cro-grinding tools for finishing metallic surfaces euspen’s 22 nd ICE Geneva Switzerland [7] Park J, Kim D, Kim H, Park W I, Lee J, Chung W 2022 Superhydrophobic Electrodeposited Copper Surface for Robust Condensation Heat Transfer ACS Omega 7 19021- 19029 [8] Xiao F, Shaojun Y, Liang B, Li G, Pehkonen S O, Zhang T 2015 Superhydrophobic CuO nanoneedle-covered copper surfaces for anticorrosion J. Mater. Chem. A 3 4374- 4388 [9] Wu X, Bai H, Zhang J, Chen F, Shi G 2005 Copper Hydroxide Nanoneedle and Nanotube Arrays Fabricated by Anodization of Copper J. Phys. Chem. B 109 22836- 22842 Aus Wissenschaft und Forschung 10 Tribologie + Schmierungstechnik · 70. Jahrgang · 1/ 2023 DOI 10.24053/ TuS-2023-0002 [3] Kuyucak S, Sahoo M 1996 A Review of the Machinability of Copper-Base Alloys Canadian Metallurgical Quarterly 35 1-15 [4] Heintze S D, Forjanic M and Rousson V 2006 Surface roughness and gloss of dental materials as a function of force and polishing time in vitro Dental Materials 22 146- 165 [5] Asadi E, Hockauf R, Denkena B, Grove T and Wurz M C 2018 Flexible monoand multilayer micro grinding tools for ultra-high precision processing and micro machining euspen’s 18 th ICE Venice Italy [6] Steinhoff L, Dencker F and Wurz M C 2022 Production and characterisation of batch manufactured flexible mi-