Journal of Micromechanics and Microengineering
`
`Electrodeposited copper inductors for intraocular
`pressure telemetry
`
`To cite this article: R Puers et al 2000 J. Micromech. Microeng. 10 124
`
`View the article online for updates and enhancements.
`
`Related content
`
`-
`
`-
`
`-
`
`Topical Review
`Rogier A M Receveur, Fred W Lindemans
`and Nicolaas F de Rooij
`
`Focused ion beam applications in
`microsystem technology
`Steve Reyntjens and Robert Puers
`
`Effects of direct and pulse current on
`copper electrodeposition
`throughphotoresist molds
`Jean-Marie Quemper, Elisabeth Dufour-
`Gergam, Nadège Frantz-Rodriguez et al.
`
`Recent citations
`
`-
`
`-
`
`-
`
`Wireless pressure sensor integrated with a
`3D printed polymer stent for smart health
`monitoring
`Jongsung Park et al
`
`A Subcubic Millimeter Wireless
`Implantable Intraocular Pressure Monitor
`Microsystem
`Hansraj Bhamra et al
`
`Minimizing a Wireless Passive LC-Tank
`Sensor to Monitor Bladder Pressure: A
`Simulation Study
`Jacob Melgaard et al
`
`This content was downloaded from IP address 72.37.250.188 on 02/04/2019 at 19:25
`
` 
`Abbott
`Exhibit 1019
`Page 001
`
`

`

`J. Micromech. Microeng. 10 (2000) 124–129. Printed in the UK
`
`PII: S0960-1317(00)09320-7
`
`Electrodeposited copper inductors for
`intraocular pressure telemetry
`
`R Puers, G Vandevoorde and D De Bruyker†
`
`Katholieke Universiteit Leuven, dept. ESAT-MICAS Kardinaal Mercierlaan 94,
`B-3001 Heverlee, Belgium
`E-mail: bob.puers@esat.kuleuven.ac.be,
`glenn.vandevoorde@esat.kuleuven.ac.be and
`dirk.debruyker@esat.kuleuven.ac.be
`
`Received 10 January 2000
`
`Abstract. A microsystem for wireless long-term measurement of the intraocular pressure is
`presented. The sensing element is a novel distributed parallel-resonant inductive–capacative
`circuit, with a pressure-dependent resonance frequency. This circuit is based upon a twofold
`on-chip deposited inductor. The high Q inductor is deposited by electrodeposition of copper
`on a micromachined chip incorporating a pressure-sensitive diaphragm. Test structures were
`fabricated and characterized. Q factors of 30 at 45 MHz and inductance values of 0.4 µH are
`obtained for 3 × 3 mm2 structures.
`(Some figures in this article are in colour only in the electronic version; see www.iop.org)
`
`glaucoma patient
`
`normal patient
`
`(+/- 500 Pa)
`daily variation
`
`3200
`
`ocular hypertension
`
`2800
`
`normal
`
`2400
`
`2000
`
`1600
`
`0
`
`Time (h)
`
`24
`
`0
`10
`Time (s)
`
`fluctuation due
`to heartbeat
`
`atmosphere)
`(relative to
`
`IOP (Pa)
`
`100 - 400 Pa
`
`Figure 1. Fluctuations of the IOP for normal and glaucoma
`patients.
`
`circuit (LC-circuit), with a pressure-sensitive resonance
`frequency. This idea was implemented by Rosengren et al
`[4] who demonstrated a micromachined capacitive pressure
`sensor placed in parallel with a hand-wound inductor coil.
`The resulting parallel-resonant LC-circuit (with pressure-
`sensitive capacitance) is read out by exciting the sensor over
`a frequency range and detecting resonance in a separate
`detection circuit (see figure 2(a)).
`The sensor chip is
`permanently implanted in the eye, without disturbing the
`normal vision. Another approach is adopted for patient cases
`where an artificial intraocular lens is implanted: both Van
`Schuylenberg et al [1] and Mokwa and Schnakenberg et al [5]
`
`1. Introduction
`
`Long-term measurement of the intraocular pressure (IOP)
`is of great
`importance in the diagnosis and treatment
`of glaucoma, a condition characterized by an anomalous
`increase of the IOP. This pressure is normally at an average
`level of about 2100 Pa above the atmospheric pressure, with
`daily cyclic variations of about 530 Pa and fluctuations of
`130–400 Pa associated with each heart beat (see figure 1).
`Persistent glaucoma can lead to axonal degeneration of the
`optic disc and to blindness [1]. The classical measurement
`technique of the IOP using external mechanical instruments
`(an art called tonometry) is unable to provide both accurate
`and long-term measurement data [2]. An ideal solution,
`responding to this last demand in particular, would consist
`of a fully implantable, telemetric pressure sensing system.
`Stringent miniaturization and bio-compatibility require-
`ments lead to a microsystem based on a micromachined
`pressure sensor as the sensing element, and using inductive
`excitation as well as detection. Our laboratory has been in-
`volved in research on this topic for several years and we
`present here some aspects of our work, in particular the real-
`ization of high Q planar inductors on the sensor chip. First
`we will describe our proposed approach and the actual sensor
`design.
`
`2. IOP sensor
`
`2.1. Overview
`
`An early solution was proposed by Collins [3] and
`consisted of an implantable passive inductive–capacitive
`
`† Corresponding author.
`
`0960-1317/00/020124+06$30.00 © 2000 IOP Publishing Ltd
`
`Abbott
`Exhibit 1019
`Page 002
`
`

`

`1L
`
`Excitation
`
`K 1
`
`Sensor
`
`K 2
`
`L
`
`C
`
`Detection
`
`L3
`external unit
`
`K 3
`link
`
`(a)
`
`sensor chip
`implanted in eye
`
`D
`
`Sensor chip
`(b)
`
`L2N+1
`
`L2N-1
`
`L3
`
`CN
`
`CN-1
`
`C2
`
`C1
`
`L1
`
`L2N
`
`L4
`
`L2
`
`(c)
`
`Figure 2. Schematic diagrams (a) of parallel-resonant IOP sensor.
`The excitation coil is swept over a frequency band containing the
`resonance frequency of the LC-circuit. The detection coil
`measures the resonance peak. The inductive links are indicated.
`(b) Novel approach: distributed resonant LC structure combined
`with a nonlinear element (diode) for the generation of second
`harmonics. (c) Electrical equivalent model of the distributed
`structure (the resistive elements are not shown).
`
`have demonstrated the integration of a telemetric tonometer
`system into the artificial lens.
`
`2.2. Sensor design
`
`In this paper we elaborate on a similar concept to that
`discussed by Rosengren et al
`[4]:
`the realization of
`an implantable sensor chip with wireless excitation and
`detection. Our approach, however, contrasts with the
`mentioned solution on several points:
`First, the inductor coil is integrated on the silicon sensor
`chip, hereby complying with the dimensional constraints (the
`complete system should fit in a box of 5×3×1 mm3) in a more
`optimal way than would be possible using a discrete inductor.
`At the same time, some difficult packaging issues resulting
`from the use of such a discrete component are avoided.
`Second, due to the small size, the resulting inductive
`links (k1 and k3 in figure 2(a)) are weak (the coupling factors
`are typically in the order of 1%) and the (unwanted) direct
`coupling between the drive and the detection coil will obscure
`
`Electrodeposited copper inductors for IOP telemetry
`
`the sensor signal. Therefore, a diode connected in parallel
`is added (see figure 2(b)). The nonlinear action of the diode
`results in the generation of higher harmonics, and these can
`be more accurately measured at the detection coil. Since
`the amplitude of (for instance) the second harmonics is
`maximal at resonance, the resonance frequency (and hence
`the pressure) can be determined using this principle.
`Finally, instead of implementing a distinct inductor and
`capacitor (the latter influenced by pressure), a design is made
`of a distributed pressure-sensitive parallel-resonant circuit
`(represented schematically by figure 2(b)), which has the
`advantage of providing a higher degree of miniaturization.
`A solution using a separate inductor and capacitor would
`obviously require more chip area.
`It reveals
`Figure 3 presents the proposed sensor.
`several views of the device. The inductor coil is split
`into two parts, one half is placed on a movable diaphragm
`(micromachined in silicon), the other on a fixed substrate.
`Both chips are bonded together, leaving a small gap between
`the inductor parts. Feedthrough contacts connect both parts
`of the inductor. These are aligned in such way that the
`conducted current in both parts contributes co-operatively
`to the magnetic flux (and inductance) (see figure 3). On
`the other hand, a strong capacitive coupling between the
`two inductor parts is accomplished, and this coupling is
`dependent on pressure: due to the deformation of the
`diaphragm, the distance between the two (parallel) coil parts
`varies. The resulting ‘self-capacitance’ can be seen in first
`order as a parallel capacitor (leading again to the equivalent
`circuit of figure 2(a)), but it is, essentially a distributed
`parameter, comparable to the parasitic capacitance present
`in any inductor coil (see also [6]). One consequence of
`√
`this is that instead of one clear (parallel) resonance peak
`LC), other resonance peaks emanate at higher
`(at 1/2π
`frequencies (as shown qualitatively in the Bode diagram of
`figure 4). These secondary peaks are typical for a system of
`a distributed nature.
`A modified electrical equivalent model of the sensor is
`given in figure 2(c), with N denoting the number of turns
`of one inductor part. The resistive contribution of the metal
`lines is not represented in this figure. Note that, also, the
`inductance of the coil changes with pressure variation, but this
`effect proves to be small when compared to the capacitance
`change.
`The design, modelling and (SPICE) verification of this
`device have been performed during the course of a Master’s
`theses [7] in our group. In the rest of this article, the attention
`will be focused on inductor coil realization.
`
`2.3. Fabrication technology
`
`The parasitic resistance of the inductor has not been discussed
`yet. This will be, however, a limiting factor on the quality
`factors of the inductors and the performance of the overall
`system. Obviously, the resistance of the sensor inductor coil
`should be minimized. The sensor is to be fabricated in a
`planar lithographic process incorporating micromachining
`steps.
`Low-resistance (thus high Q) inductors can be
`achieved by depositing thick metal films with low specific
`resistivity.
`Selective electrodeposition of copper, using
`
`125
`
`Abbott
`Exhibit 1019
`Page 003
`
`

`

`R Puers et al
`
`Top Si wafer
`Pressure sensitive
`diaphragm
`
`Diaphragm with mesa
`
`4 mm
`
`4 mm
`
`Twofold copper inductor
`
`contact
`
`direction
`of current
`
`0.7 mm
`
`Cu inductor
`
`Bottom Si wafer
`
`Inner coil contact
`
`Figure 3. The proposed IOP sensor. The device consists of a stack of two silicon chips. One contains a micromachined pressure sensitive
`diaphragm. The twofold copper inductor is the sensing element.
`
`Table 1. Typical inductor design parameter values.
`
`Parameter
`
`Value
`
`Parameter
`
`Value
`
`Inductance (L)
`Capacitance (C)
`Resonant frequency (fres)
`Resistance (R)
`Qmax (= ωres L/R)
`
`0.55 µH
`23 pF
`45 MHz
`1.2
`+/ − 100
`
`Coil length
`Coil width
`Line width
`Line height
`Spacing
`
`4.1 mm
`2.1 mm
`130 µm
`12 µm
`5 µm
`
`conventional UV lithography and enables the development
`of relatively high-aspect ratio structures. The following steps
`for the realization of the test inductor coils are performed:
`• growth of 600 nm insulating SiO2 on the silicon;
`• sputtering of a 50 nm/100 nm Ti/Cu seed layer (necessary
`for the electrodeposition step, the Ti layer serves as
`adhesive layer);
`• spinning, softbake, illumination and development of
`a 12µm thick AZ4562 resist using the test mask,
`hardbake;
`• copper electrodeposition (filling the resist patterns to
`9–10 µm);
`• removal of resist and etching of seed and adhesive layers.
`Figure 5 shows a diagram of a developed resist pattern
`and a micrograph of a profile. A thickness of 12 µm and
`aspect ratio (when defined as the ratio of thickness to minimal
`line width) of two to three is obtained. Figure 6 summarizes
`the process flow.
`
`3.2. Copper electrodeposition
`
`A simple electrodeposition set-up is used (the silicon wafer
`is placed in a teflon holder, contacted by a glued copper strip
`and aligned to a parallel copper anode). The electrolyte was
`chosen such that neither the seed layer nor the photoresist
`mask would be attacked. A CuSO4.5H2O/K2SO4/HCOOH
`solution with a pH of 4 turned out to be ideal. The deposited
`copper mass is given in function of the electrodeposition
`current and time using Faraday’s law. Controlling the copper
`
`10 8
`10 7
`10 6
`F requency (H z)
`
`10 9
`
`10 4
`
`10 3
`
`10 2
`
`10 1
`
`10 0
`105
`
`Z ( Ω )
`
`Figure 4. Bode plot (amplitude) of the distributed LC structure
`(measured on a macro-model).
`
`a thick photoresist as a mask, proves to be the ideal
`candidate for this. Copper is a better conductor than
`most metals (it is only beaten by pure silver) and can
`be deposited in a controlled way using electrodeposition.
`Both electrodeposition using an external bias voltage and
`without bias (‘electroless plating’) are widely applied in the
`fabrication of PCBs [8]. Copper electrodeposition has even
`become a hot topic in the ULSI industry, since its utilization
`combined with subsequent chemical mechanical polishing
`(CMP) steps has proven its merits in realizing high-quality,
`multilevel interconnections [9].
`The next section discusses in more detail the process we
`used. Table 1, finally, gives some typical values of design
`parameters of a practical inductor coil design.
`
`3. Inductor fabrication
`
`3.1. Lithography
`
`As seen in table 1, a metal thickness of about 10 µm is
`required. A thick photoresist mask needs to be prepared
`on the silicon wafer for the subsequent electrodeposition
`of copper. The AZ4562 resist of Hoechst is suitable for
`
`126
`
`Abbott
`Exhibit 1019
`Page 004
`
`

`

`Electrodeposited copper inductors for IOP telemetry
`
`SiO2
`
`Resist
`
`A
`
`d
`
`10 µm
`
`h=12 µm
`
`resist
`profile
`
`A
`
`d
`
`FIB cut
`SiO2
`
`Resist
`
`Figure 5. Thick resist pattern. The profile on the right-hand side is obtained after a focused ion beam (FIB) cut.
`
`SiO2
`Si
`50/100 nm Ti/Cu
`
`14 um AZ4562 resist
`
`electroplated Cu
`
`seed layer stripped
`
`Figure 6. Process flow of the inductor fabrication.
`
`thickness thus requires knowledge of the ‘active area’ (the
`wafer area covered by a conductive seed layer and exposed
`to the electrolyte), which can be calculated from the mask
`patterns. Figure 7 gives some SEM micrographs of the
`deposited inductor coils (after removal of the resist mask).
`The uniformity of the deposited copper was investigated
`and the following conclusions can be drawn: non-uniform
`
`local
`material deposition is mainly caused by the
`non-uniformity of the electrical current distribution during
`the electrodeposition [10]. Non-uniform ‘current lines’
`originate from the presence of non-conducting surfaces, such
`as the photoresist masked area, in the vicinity of conducting
`surfaces. Non-uniformity in our case could be identified on
`three levels: on the wafer level, the die level and the individual
`conductor lines. In practice, more copper was deposited; (a)
`at the edges of the wafer, (b) at the outer windings of the
`inductors and (c) at the edges of the conductor lines. The
`deviation is always in the order of 1 µm. This means that
`in order to control the gap between the two inductor parts,
`an additional polishing step is mandatory. Figure 8 shows a
`typical profile.
`
`4. Measurements
`
`A set of inductor coil designs were made and implemented
`in the test mask.
`Inductance and resistance values
`of one ‘half’ of the final structure of figure 3 (planar
`rectangular and octagonal coils) were measured using
`medium-frequency impedance measurement apparatus and
`
`Figure 7. SEM micrographs of the deposited copper coils.
`
`127
`
`Abbott
`Exhibit 1019
`Page 005
`
`

`

`R Puers et al
`
`Figure 8. Dektak profile of the inductor demonstrating electrodeposition non-uniformity.
`
`Figure 9. Micrograph of a flip-chip bonding experiment using two inductor chips.
`
`Table 2. Measured and calculated L and R values for selected coil
`designs.
`
`Design type
`
`1
`2
`3
`4
`
`Lcalc
`(µH)
`
`0.50
`0.24
`0.35
`0.17
`
`Rcalc
`( )
`
`4.18
`1.59
`2.64
`2.06
`
`Lmeas
`(µH)
`
`Rmeas
`( )
`
`Qext r
`
`0.40
`0.21
`0.27
`0.14
`
`4.35
`1.63
`2.57
`1.96
`
`26
`36
`30
`20
`
`compared to calculations. These were based on the semi-
`analytical model of Crols et al [11]. A summary of both
`the experimental and calculated data for some coil designs is
`given in table 2.
`
`(by flip-chip bonding) are candidates to realize the complete
`structure. An example of a flip-chip experiment is shown
`in figure 9, where two copper inductors (on diced chips)
`were aligned and bonded together, using gold bumps. The
`adhesion of the bumps proved to very good, better than the
`adhesion of the titanium layer supporting the coil on the
`silicon, since the coil was ripped off the chip during a pull-
`test, as can be seen on the picture.
`To conclude, it can be summarized that a first step
`towards a novel IOP telemetry microsystem was taken by
`the realization of high-Q planar inductor coils integrated on
`the sensor chip.
`
`5. Future work and conclusions
`
`Acknowledgments
`
`The next step in validating the proposed sensor design
`consists, of course, in realizing the complete structure (the
`distributed parallel-resonant LC structure) and characterizing
`it. On the technological side, polishing steps following
`the electrodeposition need to be performed in order to
`obtain a uniform, well controlled copper thickness. It was
`briefly touched on that both wafer- and chip-level bonding
`
`The authors wish to thank Rita Van Hoof and Mia Honore
`of IMEC, as well as Wouter Ruythooren and Jan Fransaer of
`the Material Sciences Department (MTM) of K U Leuven for
`their assistance and electrodeposition support. Ben Helsen
`and Raf Vandersmissen are acknowledged for the substantial
`work they have performed during the course of their Masters
`Theses.
`
`128
`
`Abbott
`Exhibit 1019
`Page 006
`
`

`

`References
`
`[1] Van Schuylenberg K, Peeters E, Puers R, Sansen W and
`Neetens A 1991 An implantable telemetric tonometer
`for direct intraocular pressure measurements Proc. 1st
`Eur. Conf. on Biomedical Engineering, (Nice)
`pp 194–5
`[2] Den Besten C 1993 Sensor systems for the measurement of
`intraocular pressure PhD Thesis T.U.Delft
`[3] Collins C 1967 Miniature passive pressure transducer for
`implanting in the eye IEEE Trans. Biomed. Eng. 14
`74–83
`[4] Rosengren L, B¨acklund Y, Sj¨ostr¨om T, H¨ok B and
`Svedbergh B 1992 A system for wireless intraocular
`pressure measurements using a silicon micromachined
`sensor J. Micromech. Microeng. 2 202–4
`[5] Mokwa W and Schnakenberg U 1998 On-chip microsystems
`for medical applications Proc. Microsystem Symp. (Delft)
`pp 69–75
`
`Electrodeposited copper inductors for IOP telemetry
`
`[6] Van Schuylenberg K 1998 Optimisation of inductive
`powering of small biotelemetry implants PhD Thesis K U
`Leuven pp 218–20
`[7] Bruggeman A and Van Regemorter J C 1994
`Inductief-capacitieve druksensor als intraoculair
`druktelemetriesysteem Master Thesis K U Leuven
`(in Dutch)
`[8] Romankiw L T 1987 Electrodeposition in the electronic
`industry 1987 Proc. Symp. on Electrodeposition
`Technology (San Diego) pp 13–41
`[9] Collins G J and Taylor T C 1999 Copper process
`characterisation Eur. Semicond. 41–50
`[10] Fransaer J, Bouet V, Celis J P, Gabrielli C, Huet F and
`Maurin G 1995 Perturbation of the flow of current to a
`disk electrode by an insulating sphere J. Electrochem. Soc.
`142 4181–9
`[11] Crols J, Kinget P, Craninckx J and Steyaert M 1996 An
`analytical model of planar inductors on lowly doped
`silicon substrates for high frequency analog design up to
`3 GHz Proc. Symp. on VLSI Circuits (Honolulu) pp 28–9
`
`129
`
`Abbott
`Exhibit 1019
`Page 007
`
`