"ROLLED-UP MICRO-SOLENOIDS AND MICRO-TRANSFORMERS"

DESCRIPTION

Field of invention

The present invention relates to the following two major fields: a. The field of micro and nano-technology and the production of integrated electrical components, namely micro- solenoids and micro-transformers, and b. The field of micro- electromechanical systems (MEMS), and more specifically, the field of magnetic MEMS.

Background A solenoid is an indispensable component in many electrical circuits, in particular those providing for radio- frequency applications. Solenoids also play an important part in the field of magnetic micro-electromechanical systems (MEMS) , where it can magnetically actuate a cantilever or a mirror [1] and sense or generate a local of magnetic field for magnetic memory read-head applications.

In integrated circuits, solenoids are usually made in the form of concentric loops on a plane [26] . This form requires large area, and therefore the resulting solenoids have relatively small magnetic flux density and low inductance. A low inductance also inhibits the performance of related transformers, as well as the performance of magnetic MEMS actuators.

A superior performance was demonstrated for three dimensional solenoids and transformers but these are complex to manufacture [26] . This impasse may be overcome using a novel class of rolled-up micro and nanotubes, as well as scrolls [3-7]. These rolled-up structures are based on a membrane comprising a number of thin layers in contact; the membrane is partially released from the substrate by means of a selective etching of an underlying sacrificial layer; the

internal strain due to the different material layers in the membrane then causes it to curve and roll up in any number of times, depending on the lattice mismatch, thickness [21] and the distance etched. Several groups have fabricated metallic coils based on these rolled-up structures. The fabricated coils can be classified in three classes, according to the method of production. First method is to lithographically pattern a free standing diagonal bi-layer stripe [5,20]. After removal of the sacrificial layer, the stripe coils-up in a direction dominated by the Young modulus, which is strongly anisotropic in the strained Si/SiGe films used in [20] .

The other two approaches consist of patterning a metallic track onto the membrane without removing it. This patterning can be done either lithographically on a semiconducting bi-layer [10], or by using a micro-contact such as PDMS stamp to pattern gold tracks on a polymer bi- layer [14]. Either way, after disposing of the underlying sacrificial layer, the strained membrane rolls-up into a scroll with the embedded metallic track winding-up around the axis of the scroll.

The problem with the above mentioned coils and with all coils based on said scrolls known in the art (except for patent application [29] to be discussed below) , is that they can not be used as inductors. An inductor needs to comprise a helical conducting path with a net number of clockwise or counter-clockwise revolutions around the axis of the scroll. But in the conducting coils mentioned above, if the metallic track is connected to the two electrodes (which are on the substrate) , an electrode at each end, the net helicity (number of revolutions around the axis of the scroll) of the coil is zero.

This problem is addressed by the patent application in reference [29] . The solution offered there is to extend the conducting track to the side edge of the membrane wherein a

liquid metal drop may contact that edge, and form the connection between the buried end of the track and an electrode or with another side track which is connected with an electrode. The cited patent application [29] also discloses transformers based on two such solenoids rolled-up in conjunction one against the other.

Although the solution offered by reference [29] allows, in principle, to produce scroll-based solenoids and transformers, it has a number of disadvantages. The main disadvantage is the need to treat the one side edge of the scroll with liquid metal drop. Such a procedure is highly unconventional and may pose difficulties in scaling and precision.

A second disadvantage lies in the low coiling density, and hence low induction, of coils produced in the cited method [29] .

Brief Description of the Invention

The prime aim of the present invention is to disclose novel forms of solenoids and transformers which solve the above mentioned problem. The secondary aim of the present invention is to disclose a number of magnetic transducers based on the disclosed solenoids.

The solenoids disclosed in the present invention have three major advantages over the similar solenoids disclosed in the cited patent application [29] ; first, the fabrication of the solenoids disclosed here is simpler, requiring the use of only standard lithographic techniques and in particular, not requiring liquid metals, and secondly, it allows higher loop density, translating into larger inductance, and thirdly, a simple modification of design provides for a transformer on a single rolled-up membrane, without requiring two membranes as in [29] .

Let a membrane consisting of two or more thin material layers lie on top of a sacrificial layer. Said membrane then

rolls-up into a scroll when said sacrificial layer is etched away (the well known technical procedures of the etching and rolling-up can be found, for example, in the references [3- 7] ) . Call the end of the membrane where the etching begins, the start edge, and where the etching stops, the stop edge. The scroll is then being formed with the start edge curving in the direction of the stop edge, and revolving around its axis several times.

Now let there be two electrodes patterned at two different places at the stop edge. A conducting track connected at one end to one electrode, begins at the stop edge and extends to the start edge consist of a helical path wherein the helicity is equal to the number of revolutions of the underlying scroll. A further extension of the track from the start edge to the second electrode makes the reverse helical path so that the net helicity is zero.

The problem is how to make a non-zero net helicity. The solution offered in the present invention is the following. Consider the path of the current along one complete loop to be comprised of two distinct paths, connected in series. The inward path, from the first electrode or, in general, from the stop edge toward the start edge, follows a helical path along a conducting track. The outward path, from the end of the inward path to the second electrode or, in general, toward the stop edge, follows a radial path. The reason for the radial path is that it carries no helicity so it does not cancel the helicity of the inward path.

In the present invention, the current is made to flow radially by designing a membrane which includes sections which are metallic through the thickness of the membrane, so that in the scroll neighboring sections lie on top of each other, allowing for the current to flow in a direction normal to the faces of said membrane.

In other words, each current loop consist of an inward, helical path and then an outward radial path; in this way the

coil in the scroll forms a solenoid with a net non-zero helicity.

The inward paths consist of the conducting tracks, are metallic stripes deposited onto (or below or within) the membrane. At any rate, said tracks extend only through part of the thickness of the membrane. The outward paths consist of the conducting sections, on the other hand, extend through the entire thickness of the unrolled membrane. Said sections can be manufactured by a variety of standard techniques. For example, first said membrane is perforated by holes etched in it. Then said holes are filled by the deposition of metallic material. After the release of the sacrificial layer, the membrane rolls-up with a somewhat increased radius of curvature compared to the un-perforated membrane. This increase of the radius, is due to the stiffness of the conducting sections, ie-the filled holes.

Brief Description of the Drawings

The present invention is further elucidated in more detail below using the following exemplary embodiments in conjunction with the drawings, in which:

- Figure Ia depicts a partial side view of the prerelease membrane,

- Figure Ib depicts a partial side view of the rolled-up membrane after the release of the sacrificial layer,

- Figure Ic depicts a top view of the pre-released membrane which forms a solenoid upon rolling-up,

- Figure 2 depicts a top view of the pre-released membrane which forms a transformer upon rolling-up,

- Figure 3a depicts a partial side view of the solenoid on a cantilever,

- Figure 3b depicts a partial top view of the solenoid on a cantilever,

- Figure 4 depicts a partial side view of two solenoid as actuating element for a mirror,

- Figure 5a depicts a partial side view of two connected membranes at the pre-release stage, - Figure 5b depicts a partial side view of said two connected membranes after the top membrane was released,

- Figure 5c depicts a partial side view of the final stage of said two connected membranes, - Figure 6 depicts a partial side view of a solenoid under a mirror,

- Figure 7a depicts a partial side view of a solenoid on a cantilever formed in a trench,

- Figure 7b depicts a partial side view of a solenoid on a cantilever with the walls of said trench etched away,

- Figure 7c depicts a plan view of Figure 7a,

- Figure 8a depicts a partial side view of a solenoid on cantilever in a magnetic environment, and - Figure 8b depicts a partial front view of a solenoid on cantilever in a magnetic field environment.

Description of the Preferred Embodiment of the Invention The principle embodiment of the invention is a solenoid. The solenoid comprising a scroll which is composed of a thin- film membrane which, after a partial release from the substrate, has rolled-up several revolutions about its axis, as already described above. Figure Ia shows a typical partial side view of the structure before the release. It comprises a substrate (1), a sacrificial layer (2), and a thin-film membrane. The membrane in figure Ia comprising two different materials (7,8), and an optional ferromagnetic layer (6). The etching of the sacrificial layer begins under point (4), which is referred to as start end, and terminates

approximately under point (3), the stop end. The details of the conducting tracks are not shown in this side view.

After removal of the sacrificial layer (2), the membrane rolls-up (see figure Ib) towards the stop end (3) due to the mismatch in the lattice constants between materials (7) and

(8). The details of material layers (7) and (7) are not shown in the side view Ib. Of course, it is not mandatory for the membrane to include only two different material layers, nor is it necessary to include the ferromagnetic layer (6) . Thin membranes of any number of different material layers are known to curve as well [12], however, the thicker the membrane, the larger is the curvature radius. Since it is generally advantageous to obtain a smaller radius of curvature, we depict in this embodiment only two layers. The radius of curvature in this case is given by the formulae in [21] . The ferromagnetic layer was added here as an optional magnetic core. That is especially beneficial in the embodiment of a transformer.

While the material layers (7,8) as well as the conducting stripes can be epitaxially grown by the standard lithographic methods applied in the semiconductor industry, the ferromagnetic layer is, in addition, usually electroplated or added in a different manner [I].

The final shape of the scroll depends on the length of the released membrane and the radius of curvature. The membrane depicted in Ib forms a scroll, after rolling-up slightly more than 3 revolutions about its annuls. Neighboring walls belonging to consecutive revolutions are attached due to van der Waals interaction (for a relevant quantitive treatment of this adhesion see [24]).

The details of the conducting tracks are exposed in the top view (figure Ic) of the pre-released solenoid membrane. Here, (1,2) are the two electrodes; the vertical line (3) is the stop end and the vertical line (4) is the start end. The ferromagnetic layer on part of the thickness is (6), and the

conducting tracks (diagonal lines) are (5) . The conducting sections are marked by crossed lines.

When the sacrificial material layer under said membrane is etched away, the start end (4) rolls-up a number of times (according to the known art) until the stop end (3), to result in a scroll (figure Ib). The distance (along the horizontal axis) between said conducting sections is approximately the circumference of curvature, so that when said membrane is rolled-up into a scroll, neighboring conducting sections lie one above the other. The conducting sections in figure Ic for example, are spaced so that in the rolled-up state, A lies on B which lies on C which lies on D.

In this way, between the inner layer of the scroll and the outer layer, the current can flow from one conducting section to another. This forms in three dimensions a radial path, which, importantly, carries no helicity.

The current in the scroll of figure 1 then flows helically inward from electrode 1 to the first conducting section, then outward radially, then inward helically to A, and outward radially to D and electrode 2. The current path then has a net helicity; in figure Ic for example, the net helicity is 3x2=6, since the membrane rolls-up three times (A on B, B on C and C on D), and there are two conducting tracks.

This arrangement achieves the desired goal of providing for a connected path for electrical current with net spatial helicity, which is required for an inductor.

The main source of loss in this configuration can be the contact resistance between neighboring conducting sections. In order to minimize this resistance the layers should be grown as smooth as possible. Following the rolling-up process, a heat treatment can improve the microscopic adhesion and reduce the resistance between neighboring sections.

A major advantage of this form of solenoids is that two different solenoids can form a transformer on a single membrane, in stark difference with the known art [29] .

The embodiment of said transformer is detailed below. A transformer comprises at least two intertwined solenoids, and preferably, a ferromagnetic core.

The side view of a transformer is therefore identical to the solenoid, as partially depicted in figure Ia (before removal of sacrificial layer) and figure Ib (after the removal) . The layout of the conducting tracks and sections of a transformer is depicted in figure 2. Here is a top view of a membrane in the unrolled state comprising electrodes 1-4 and a ferromagnetic core 6.

As in the solenoid embodiment, there are two types of conductors; those marked by crossed lines, the conducting sections, conduct through the thickness of the membrane while those marked by diagonal lines, the conducting tracks, conduct within only part of the thickness of the membrane.

In the final, rolled-up state, section A (figure 2) is attached to section B, since the distance between them equals roughly the circumference of the curvature; likewise, section B is attached to section C, section C is attached to section D and section D is attached to section E and to the second electrode. Since all said sections are metallic through the thickness of said membrane, the current path from section A to electrode 4 takes the shortest, radial trajectory. The current from A to E (and the parallel, unmarked sections) thus carries no helicity.

Hence, the current in the loop from electrode 1 to electrode 4 carries the net helicity (2x4) of the two conducting tracks (the number 4 is the number of revolution the membrane of figure 2 to undergo upon release) . As depicted in figure 2, the separated positions of the conducting sections allow for the intertwining of another solenoid, and hence, the formation a transformer. The second solenoid in the particular example in figure 2 contains a single conducting track. Upon rolling-up, said second

solenoid comprises a path between electrodes 2 and 3 and helicity of four.

The solenoid can be applied both as a stand alone inductor on a chip, and as a transducer element if its base comprises a cantilever or other moving part. Figures 3a, b depict such an embodiment. Figure 3a is a partial side view of the solenoid (1) on a cantilever (2), with a magnetic substrate (3). The top view, Fig. 3b, depicts in addition the two electrodes (4 and 5), and the two legged configuration of this particular cantilever.

In the embodiment of Fig. 3, the solenoid functions both as a magnetic actuator and as a sensor. The actuation is provided by an input of current / through the solenoid; this current provides a magnetic moment μ = InR ² N , where R is the approximate radius of the scroll and N is the number of revolutions of the current path in the scroll. If the magnetic substrate (3) provides a magnetic field B , the coupling with the magnetic moment brings about a force ofF -V(μ-B) on the cantilever. The cantilever for this solenoid can also be grown in a more vertical position inside a trench. This position may be preferred in order to reduce the overall footprint of the cantilever. The methods of fabrication a trench with a required shape and size are well known (see for example reference [I]). Once the trench is in place, the sacrificial layer and the layers of the membrane are grown on one of the side walls of the trench. This vertical growth gives of course a reduced footprint.

Figure 7a is a partial side view of a solenoid on a cantilever grown inside a trench, wherein the cantilever is still attached to the sacrificial layer. The electrodes, (1) and (2) are laid-out mutually perpendicular, appropriately for a matrix array. Magnetic material should be deposited at some vicinity to the cantilever (not shown) to generate a magnetic force when current passes in the solenoid, as

discussed above. (3) is an insulating layer, (4) is the cantilever layer before release, (5) the sacrificial layer and (6) is the substrate in the shape of a trench. (7) is the layer connected with the solenoid and (8) is (schematically) the solenoid. The electrical connections of electrodes (1) and (2) with said solenoid (8) are not shown.

Figure 7b depicts the same arrangement after the sacrificial layer and the trench where removed.

The solenoid transducer is not restricted, of course, to be mounted on top of the cantilever in the embodiment of Figure 3. It may likewise be mounted above or under any other moving element. An example of the latter is depicted schematically in Figure 4.

Figure 4 depicts a mirror (1) connected to a base (2) which hosts the two solenoids (3 and 4) . The structure is connected to the substrate and to external electrodes with the elastically pliable ribbons (7 and 8). A magnetic layer (5 and 6) provides for the external magnetic field B. The solenoids are not attached to the mirror (1) directly, since that may compromise their flatness which is critical in many optical applications.

The solenoid provided in this invention can actuate a mirror in a different manner: figure 6 depicts a partial side view of the latter. Here, (1) is a mirror, (2) is a torsional hinge, (3) and (4) are patches of magnetic material with different orientations, as indicated by the arrows, and (5) is (schematically) a solenoid. The structure holding the mirror is grown on top of the solenoid. The two magnetic patches must be deposited separately, so that the first patch is magnetized before the deposition of the second. The magnetic field used for the magnetization of the second patch has to be, of course, below the critical field which would cause a reversal of magnetization in the first patch.

In a matrix array, the solenoid (5) is wired to the XY electrodes (not shown) . Of course, each such unit must also contain the proper electronic circuit to be addressable.

Above, we discussed embodiments wherein the solenoid actuates a cantilever or a torsion-based mirror. The solenoid in this invention, which is based on a rolled-up membrane, can also be grown on another rolled-up membrane, such as depicted in figure 5. The pre-release side view in figure 5a shows a substrate (1), first sacrificial layer (2), first membrane (3), connecting pad (4), second sacrificial layer

(5), second membrane (6), and landing pad (7). The intra membrane layering is not shown. Using light sensitive etching, the light can be directed to first remove the sacrificial layer (5) (figure 5b) to form the small radius scroll (6), and subsequently the sacrificial layer (2) is removed to form the large radius structure (3) .

Now if the scroll (6) forms a solenoid (electrical contacts not shown) , the embodiment depicted in figure 5c is a magnetic sensor and actuator on a flexible cantilever, wherein said cantilever here comprises the membrane (3).

As with the other embodiments of a magnetic actuator above, a spatially varying magnetic field (not shown) is required to actuate the cantilever. Such a field can be provided either by external magnets (as in figures 3, 4 or 9) , or by pair of solenoids fixed to the substrate.

Since with a given current, the magnetic force is proportional to the gradient of the magnetic field, it is desirable that the magnetic environment around the solenoid- actuated cantilever posses the largest possible field gradient. Figure 9 partially depicts an embodiment comprising a large magnetic field gradient. In figure 8a we depict a partial side view comprising (schematically) a cantilever (1) and a solenoid (2) with two magnetic layers (3) and (5), a thin (optional) non-magnetic layer (4) and the substrate (6). Like in the above mentioned embodiment of a magnetically

actuated mirror (figure 6), here too, the two magnetic layers have opposite polarity, in order to generate a large change in the magnetic field in its vicinity. The different orientations are marked here by the head and the tail of an arrow.

Figure 8b shows a front view wherein the cantilever and solenoid (2) move vertically in the gap between the magnetic layers. In this view additional magnetic layers (7) and (9) are depicted, as well as an additional (optional) non- magnetic layer (8). The magnetic orientation is marked by the arrows .

The magnetically actuated cantilevers disclosed in the above embodiments convert current to mechanical motion, via the magnetic field generated by the inductor. The other way around, wherein the cantilever acts as a sensor, is equally applicable. The above embodiments double as sensors by converting mechanical motion to current, via the electromotive force generated in the inductor as the flux through it changes during the motion of the cantilever. In order to generate large electromotive forces in the inductor from the relatively small motion of the cantilever, the environment should have a large gradient of the magnetic field. The cantilever in figure 9 for example (already discussed above as an actuator) , can be configured as an accelerator sensor, wherein an acceleration moves the solenoid (2) on the cantilever through a changing magnetic field, generating a current through the solenoid.

Numerous detectors and sensors in the art are configured to detect a change in the resonance frequency of a cantilever [1] ; such a change signals the adhesion of a particle of a particular mass on the cantilever. In the present invention, the cantilever generates an oscillatory current in the solenoid. A change in the resonance frequency of the cantilever or in the shape of the oscillatory current can be used as means of sensing the adhesion of particles.

An additional functionality can be extracted from the above embodiments of a solenoid on cantilever, if the cantilever's landing-pad is made ^λ sticky' . The property of two surfaces to 'stick', or adhese, is the consequence of the van der waals-Casimir interaction, and can be tuned by an appropriate surface treatment [24].

Now consider the situation in which (see figure 9, for example) the current through the solenoid pulled the cantilever's free end to touch the substrate (not shown), or landing-pad (see (7) in figure 5c). If the adhesion force is larger than the cantilever's restoring force, the cantilever gets immobilized to the substrate. Depending on the balance of these two forces (a long range restoring force and a short range contact force) , an additional increment of force is required to liberate the cantilever. Since these forces are well known and can be tuned using standard MEMS manufacturing techniques, it can be configured so that only a small force is required to liberate the cantilever from the landing-pad. In this case, a large displacement, as well as a large amplitude vibration, is unleashed by a relatively small applied force.

The source of the large displacement and large amplitude vibration is, of course, the elastic energy stored earlier by the adhesion of the cantilever to the landing-pad. Now since part of the vibration energy is converted to current through the solenoid, the embodiment (figure 5 or 9) can function as an energy storage element. This element stores elastic energy by pulling the cantilever to the landing-pad, and extract some of this energy back, by detaching the two (with relatively small force), and extracting the current due to the resulting vibrations .

Without additional augmentation however, a large array of these solenoid-on-cantilever elements may not generate a net current, from the following reason. The extracted current from a single solenoid will have the shape of a damped

harmonic oscillator, that is: a decaying AC current. A superposition many such elements, all out-of-phase with each other, combines to zero. Therefore, in order to avoid this cancellation, the current extracted from each solenoid-on cantilever element must pass through an AC-to-DC converter, before the currents from the other elements of the array are joined. Such converters are of course well known in the art.

10

15

20

^' 25

30

35

REFERENCES:

I. Chang Liu, Foundations of MEMS, Pearson Prentice Hall, New Jersey USA 2006. 2. R. People and J. C. Bean, Appl . Phys . Lett. 47, 322 (1985) .

3. V. Ya. Prinz et al . , Free standing and overgrown InGaAs/GaAs nanotubes, nanohelices and their arrays, Physica E 6, 823 (2000) . 4. A. V. Prinz, V. A, Prinz, V. A. Seleznev, Semiconductor micro- and nanoneedles for microinjections and ink-jet printing, Microelectronic engineering 67, 782 (2003) .

5. V. Ya. Prinz, A new concept in fabricating building blocks for nanoelectronic and nanomechanic devices,

Microelectronic engineering 69, 466 (2003) . (Review article) .

6. S. V. Golod et al., Freestanding SiGe/Si/Cr and SiGe/Si/SiN/Cr microtubes, Appl. Phys. Lett. 84, 3391 (2004).

7. Oliver G. Schmidt, Kart Eberl, Thin solid films roll up into nanotubes, Nature 410, 168 (2001) .

8. O. G. Schmidt, N. Y. Jin-Phillipp, Free-standing SiGe-based nanopipelines on Si (001) substrates, Appl. Phys. Lett. 78, 3310 (2001).

9. Ch. Deneke, O. G. Schmidt, Real-time formation, accurate positioning and fluid filling of single rolled- up nanotubes, Appl. Phys. Lett. 85, 2914 (2004).

10. O. G. Schmidt et al., Self assembled nanoholes, lateral quantum-dot molecules, and rolled-up nanotubes,

IEEE journal of selected topics in quantum electronics 8, 1025 (2002) .

II. 0. G. Schmidt et al . , Free standing semiconductor micro- and nano-objects, Materials science and engineering C 19, 393 (2002) .

12. Ch. Deneke et al., Radial superlattices and single nanoreactors, Appl . Phys . Lett. 84, 4475 (2004).

13. R. Songmuang et al., Rolled-up micro- and nanotubes from single-material thin films, Appl. Phys. Lett. 89, 223109 (2006) .

14. V. Luchnikov et al., Self-rolled polymer and composite polymer/metal micro- and nanotubes with patterned inner walls, Advanced materials (2006) . 15. P. O. Vaccarro et al., Strain-driven self- positioning of micromachined structures, Appl. Phys. Lett. 78, 2852 (2001) .

16. Jose M. Zanardi Ocampo et al., Optical actuation of micromirrors fabricated by the micro-origami technique, Appl. Phys. Lett. 83, 3647 (2003).

17. Y. Takagaki et al . , Strain relaxation in AIN/GaN bilayer films grow on gamma-LiA102 (100) for nanoelectromechanical systems, Appl. Phys. Lett. 84, 4756 (2004). 18. O. Schumacher et al., Lithographically defined metal-semiconductor-hybrid nanoscrolls, Appl. Phys. Lett. 86, 143109 (2005) .

19. W. J. Arora et al . , Membrane folding to achieve three-dimensional nanostructures : Nanopatterned silicon nitride folded with stressed chromium hinges, Appl. Phys. Lett. 88, 053108 (2006).

20. L. Zhang et al., Anomalous coiling of SiGe/Si and SiGe/Si/Cr helical nanobelts, Nano letters 6, 1311 (2006) . 21. M. Grundmann, Nanoscroll formation from strained layer heterostructures, Appl. Phys. Lett. 83, 2444 (2003) .

22. Flexible electrostatic actuator; Patent number: WO2005104717.

23. S. Mendach et al., Light emission and wave-guiding of quantum dots in tube, Appl . Phys . Lett. 88, 111120 (2006)

24. F. W. Delrio et al., The role of van der Waals forces in adhesion of micromachined surfaces, Nature

Materials 4, 629 (2005)

25. H.J. Cho and C. H.Ann, A bidirectional magnetic microactuator using electroplated permanent magnet arrays, Journal of microelectromechanical systems 11, 78 (2002)

26. D. Niarchos, Magnetic MEMS: key issues and some applications, Sensors and actuators A, 109, 166(2003).

27. U. Srinivasan et al., Alkyltrichlorosilane-based self-assembled monolayer films for stiction reduction in silicon micromachines, Journal of microelectromechanical systems I ₁ 252 (1998) .

28. T. Abe et al., Effects of elevated temperature treatments in microstructure release procedures, Journal of microelectromechanical systems 4, 66 (1995). 29. O.G.Schmidt and C. Deneke, Production of microscopic and nanoscopic coils, transformers and capacitors by rolling or folding over conductive layers during removal of auxiliary layers from a substrate; USA patent application number: 20050118733.