next up previous contents
Next: Results Up: A single magnetic particle Previous: Results and discussion   Contents


Sample preparation

The complete sample preparation for the TMR elements and the manipulation system on top is a very elaborate process. It involves four different lithography steps and takes about 10 days altogether. This section gives a detailed description of all preparation steps.

The first preparation step is the sputtering of the TMR layer stack. While the development of the stack was described before, there are two things that are vital for the success of sputtering a working TMR layer system. The size of the Si-wafer substrate has to be 15$\times $15mm, because this is the minimum size for both optical lithography steps, and it is the maximum size for the sputtering in a magnetic mask. It is also vital that the substrate is electrically connected to the substrate holder with silver paste. Otherwise, the tunnel barrier is broken. The TMR layer stack is sputtered in the professional sputtering machine CLAB 600 from LEYBOLD (confer section 2.1).

Starting with the Si-wafer and the TMR layer stack on top, the samples undergo many steps until the final structure is achieved. Figure 6.5 illustrates all necessary steps of the structuring process. As first lithographic step, the supply line for the bottom contact is structured with e-beam lithography directly in the center of the sample surface (a) (please see section 2.4 for more information on e-beam lithography). This structure is etched for 1750secs (see section 2.2 for more information about the etching process) into the layer system (b) before the resist is removed (c) and the second e-beam lithography step is performed. Now, the design for the TMR elements is written to the e-beam resist (d). This and all subsequent lithography steps have to be aligned to the first structure. To do this more easily, a special cross like structure is written during the first lithographic step in the center of the design. After the development of the resist, the sample is etched again. The etching time is now 2000secs, so only the bottom contact lines and the elements remain (e). Before the resist is removed, the elements are covered with an insulating layer of 100nm SiO$_2$ (f). Removing the resist leaves the protected TMR elements with an unprotected top contact (g). The bottom contact of the element is contacted through a short-circuited TMR element (see right element in the sketches).

Figure 6.5: Side view of all preparation steps for the TMR elements and the manipulation system on top.
[]\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep1}         []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep2} []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep3}         []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep4} []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep5}         []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep6}

[]\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep7}         []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep8} []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep9}         []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep10} []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep11}         []\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep12}

To structure the contact lines, optical lithography is used because it is faster for big structures. For the optical lithography, the area is exposed where the resist is removed during the development (h), because a positive resist is used (please see section 2.3 for more information about the optical lithography). A layer of 50nm gold is sputtered as contact lines for the TMR elements (i). Below and above the gold layer, tantalum is used as an adhesive agent between gold and glass. The tantalum is essential, because without the adhesive agent the protection layer is not completely sealed, and so the TMR elements can easily be destroyed. After removing the resist, the structuring process for the TMR elements is finished. Then, all elements are measured and checked.

Only if the elements are fine, another protective SiO$_2$ layer of 100nm is sputtered in a mask that keeps the contact pads free (j). With the final lithography step (k), the design for the positioning system is structured on top of the TMR elements (chapter 5 describes in detail the development of the positioning system). 100nm gold is sputtered into the structured positioning design (l), again using tantalum as adhesive agent between SiO$_2$ and gold. After removing the resist, the sample is finished (m).

The TMR element on the left can be contacted by the top gold contact to the left and the bottom contact through the short-circuited element on the right. Besides the supply lines, it is completely enclosed in insulating SiO$_2$. The top view of the completed sample (n) reveals how the positioning system for the magnetic markers is set on top of the elements. Exact alignment is again mandatory for successful positioning.

Finally, the completed sample is cut in smaller pieces to fit into a IC-socket. It is glued into the socket with silver paste and connected by wire-bonding. Thus, the sample can be measured easily in the given setup (confer section 2.7 for more information about the measuring system).

[Side view of the completed sample.]\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep13}         [Top view sketch of the finished sample. The red dashed line shows the cut of the side view.]\includegraphics[width=.4\textwidth]{Bilder/TMR-Prep14}


For these samples, the conducting lines of the positioning system are not embedded into the underlying SiO$_2$ layer, as it was done in the chapters before. This was necessary because we want to maximise the magnetic stray field of the markers, and to achieve this, the distance between marker and sensor should be as small as possible. The protection layer between the conducting line for the positioning system and the water is not mandatory, and so this setup is chosen. The magnetic marker can be positioned in the corner, right beside the conducting line, so the distance between marker and sensor is only about 150nm (100nm SiO$_2$ protection layer and 50nm gold from the top contact line).


next up previous contents
Next: Results Up: A single magnetic particle Previous: Results and discussion   Contents
2005-07-23