8/2/2019 bosco - 2010 - mee

1/4

Self-aligned cantilever positioning for on-substrate measurements using DVD

pickup head

F.G. Bosco a,*, E.-T. Hwu b, S. Keller a, A. Greve a, A. Boisen a

a Department of Micro- and Nanotechnology, Technical University of Denmark, Lyngby 2800, Denmarkb Institute of Physics, Academia Sinica, Nankang, Taipei 11529, Taiwan

a r t i c l e i n f o

Article history:

Received 14 September 2009Received in revised form 11 December 2009Accepted 15 December 2009Available online 23 December 2009

Keywords:

Cantilever-based sensingMicrofabricationSU-8DVD pickup headChip holder

a b s t r a c t

In this paper, we present a novel approach for measuring the resonant frequency of cantilevers fabricatedin polymeric materials. We re-designed the use of a commercial DVD-ROM pickup head and combine itwith a glasspolymer substrate in order to obtain a light and portable device to measure the resonantfrequency of polymer cantilevers. The use of the Pyrex-SU-8 clamping substrate allows an easy replace-ment of the cantilever chips and a fast alignment process to the DVD-ROM laser beam. We show mea-surements of thermal noise for SU-8 and TOPAS cantilevers in air and liquid environment.

2009 Elsevier B.V. All rights reserved.

1. Introduction

Cantilever-based sensors are promising miniaturized sensingtools for bio-chemical applications [1]. The mechanical responsecan be acquired through an optical setup, where a laser beam is fo-cused and collimated on the cantilever tip, and the reflected light iscollected by a photodetector. These types of sensors can be used forbio-chemical detection when the cantilever is functionalizedwith asensing layer that interacts with the target biomolecules. The sen-sitivity of the device is thus related both to the elasto-mechanicalproperties of the cantilever and to the opto-electronic characteristicof the readout setup.

At present the optical equipment used to sense the change inthe resonant frequency due to the selective binding of biomole-cules to the cantilever surface is typically big and bulky because

precise and delicate laser alignment is required. In traditional ap-proaches a single chip is mounted on a holder, precisely alignedto the laser beam, and the cantilevers resonant frequency is mea-sured through the oscillating position of the laser spot onto thephotodetector. Typical alignment processes can take several min-utes of experienced human work. In this paper we present a novelapproach to measure the change of resonant frequency of cantile-vers by using a light, compact and high throughput optical device,described in Section 2. The distinctive feature is the auto-align-

ment of the laser beam to the cantilever tip, facilitating fasterand more precise measurements.

The optical readout of cantilever-based sensors has been re-de-signed and optimized combining the technology of commercialDVD-ROM readers [2] with a holding substrate. The holding sub-strate consist of a Pyrex support with an SU-8 structure on the top.

2. Setup

In our system cantilever chips are clamped on a predefined SU-8 structured holder while the DVD-ROM reader is placed 1 mm be-low the Pyrex substrate.

Up to 100 chips can be clamped on the substrate. Once the sub-strate is mounted onto a XY-stage, simply placing the DVD pickuphead 1 mm below the Pyrex surface will automatically align 800cantilevers to the optical system, being each cantilever tip placedat the same Z distance from the optical head. This distinctive fea-ture saves several hours of alignment work when sequential mea-surements over several chips need to be performed, compared toexisting optical setup.

As a consequence, the holding substrate plays an important rolein the setup. Chips need to be perfectly clamped to the substrate inorder to keep precisely the position of the cantilevers in the XYZspace; furthermore the cantilevers need to be oriented parallel tothe Pyrex surface.

Fig. 1 shows schematically the layout of the experimental setup.Light (650 nm) emitted by a photodiode passes through a beamsplitter and is collimated and focused by an optical system com-

0167-9317/$ - see front matter 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.mee.2009.12.064

* Corresponding author.E-mail address: filippo.bosco@nanotech.dtu.dk (F.G. Bosco).

Microelectronic Engineering 87 (2010) 708711

Contents lists available at ScienceDirect

Microelectronic Engineering

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / m e e

http://dx.doi.org/10.1016/j.mee.2009.12.064mailto:filippo.bosco@nanotech.dtu.dkhttp://www.sciencedirect.com/science/journal/01679317http://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.sciencedirect.com/science/journal/01679317mailto:filippo.bosco@nanotech.dtu.dkhttp://dx.doi.org/10.1016/j.mee.2009.12.064

8/2/2019 bosco - 2010 - mee

2/4

posed of two lenses. The laser thus passes through the 500 lm Pyr-ex substrate (n = 1.47) and is focused on the cantilever tip with a0.75 lm spot diameter. Complete gold coating or gold pads onthe bottom surface of the cantilever allow the laser beam to be re-flected back to the detection system, composed by a 4-quadrant

photodetector.Using astigmatism-based readout, the signal is obtained whensome asymmetry is present in the optical beam spot focused onthe photodetector. The principle of the astigmatic method is basedupon an optical aberration, called astigmatism, occurring when anoptical system is not symmetric about the optical axis, e.g., cylin-drical lenses [3]. In this case, any small deflection of the cantileveris converted to a small focus error onto the photodetector.

The measurement signal from the preamplifier is recorded by aPC through a data acquisition card (DAQ) which has a bandwidthof20 MHz and a resolution of 14-bit [4].

3. Substrate fabrication

One of the main challenges is the microfabrication of efficientclamping structures that allow precise chip fixation and cantileveralignment over a Pyrex aperture for the bottom-up laser reading.Therefore, several footprint shapes were fabricated in order to opti-mize the holding properties for different cantilever chip: clampingof SU-8 [5], silicon and TOPAS [6] cantilever-based chips was pro-ven. The cantilever suspending configuration is obtained by spin-coating and structuring two layers of SU-8 on a transparent Pyrexwafer (see Fig. 1). The SU-8 layers have thicknesses of 50 and300lm, respectively.

We evaluate empirically the best holding configuration of afootprint based on the ability of the clamping structure to: alloweasy installation and replacement of the chip into the holdingstructure; clamp the body chip without deforming significantly

the chip structure itself; keep the cantilever array oriented parallelto the Pyrex window. Several experiments were performed and the

resulting holder-chip structures were analyzed with optical micro-scope and SEM.

We have observed that the substrate holding properties arehighly increased moving from the traditional holder structure,where the body chip is encapsulated into a square footprint(edgeedge clamping), to configurations where the chip cornersare clamped by the footprint walls (corneredge clamping) orwhere the chip sides are clamped by concave footprint structures(edgecorner clamping) (Fig. 2).

These geometries allow better fixation of the SU-8 chips thatshow intrinsic fluctuations of the body chip dimensions due tothe photolithographic process (up to 40 lm). SU-8 corners are softenough to be slightly deformed to adapt chips fabricated with dif-ferentresolutions to precisely fit in the same footprint. Fig. 3 showsa SEM picture of the cantilever array oriented parallel to the Pyrexaperture.

Fig. 1. Laser path scheme: the laser beam passes through the Pyrex window beforebeing focused on the cantilever. A perfectly focused beam has a spot size of0.75 lm. The inset shows the asymmetric light intensity on the 4-quadarantphotodetector due to the astigmatic aberration induced by the cylindrical lens. Thefocus error signal is given by the intensity differences of the single photosensitivefragments composing the whole detector.

Fig. 2. SEM picture of the footprint configuration for SU-8 chips with best holdingproperties. It integrates both edgecorner (round markers) and corneredge(square markers) clamping points. The body chip is 2.5 4.2 mm.

Fig. 3. SEMpicture of the cantilever array oriented parallel to the Pyrex window, for

the footprint configuration shown in Fig. 2. The SU-8 cantilevers are 500 lm long,100lm wide and 5 lm thick.

F.G. Bosco et al./ Microelectronic Engineering 87 (2010) 708711 709

http://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/meehttp://www.elsevier.com/locate/mee

8/2/2019 bosco - 2010 - mee

3/4

4. Measurements

When a cantilever sensor operates in dynamic mode, the reso-nance frequency, f0, of the cantilever is monitored and as mass ad-sorb onto the structure, the resonance frequency decreases. Thischange in resonance frequency, Df, for a homogeneously distrib-uted adsorbed mass on the cantilever surface is given by:

Df f0Dm

2m01

whereDm is the mass of the adsorbents and m0 is the initial mass ofthe cantilever [5]. Thus the resonant frequency f0 is a determinantparameter to link given shift of frequency, Df, to a correspondingmass change,Dm. Here we show resonant frequency measurementsperformed on different SU-8 and TOPAS cantilevers acquired withthe re-designed optical setup illustrated in Fig. 1. In Fig. 4 resonantfrequency measurement of Au coated SU-8 cantilevers havinglength of 500lm, width of 100 lm and thickness of 5.5 lm isshown. The measurement is acquired in air. Similar measurementshave been done for TOPAS cantilevers, and in both cases the inten-sity of the reflected light is sufficient for the resonant peaks to bedetected.

Most measurements of bio/chemical reactions are performed in

liquid. Therefore the measurements have been repeated in a buffersolution. These measurements were conducted by applying a sin-

gle drop in the SU-8 defined pool. Table 1 presents average valuesof resonant frequency and Qfactor for SU-8 and TOPAS cantileversmeasured in air and liquid environment. The values of the Qfactormeasured in air are similar to the values found in literature forpolymer cantilevers. As expected we have observed a decrease ofthe value of Q for experiments conducted in liquid solution [7].

5. Conclusions

The integration of the DVD reader with the on-substrate hold-ing approach leads to a high throughput flexible platform witheasy self-alignment and replacement of the cantilever chips. Withthis new on-substrate approach tens of chips can be placed on thePyrex-SU-8 wafer and be read sequentially. Resonant frequencymeasurements in buffer solution and air were performed for SU-8 and TOPAS cantilevers, giving promising future opportunitiesfor high resolution sensing in air and liquid environment.

References

[1] M. lvarez, A. Calle, J. Tamayo, L.M. Lechuga, A. Abad, A. Montoya, Biosens.Bioelectron. 18 (2003) 649.

[2] E.-T. Hwu, K.-Y. Huang, S.-K. Hung, I.-S. Hwang, Jpn. J. Appl. Phys. 45 (3B) (2005)

23682371.[3] B. Hnilicka, A. Voda, H.J. Schroder, Sens. Actuators A-Phys. 120 (2) (2005) 49.

Fig. 4. Resonant frequency peak at 6.671 kHz measured for a gold coated SU-8 cantilever.

Table 1

Geometrical parameters, resonant frequency and Q factor for SU-8 and TOPAS cantilevers measured both in air and liquid.

l (lm) W (lm) t (lm) f air (kHz) f buffer (kHz) Q factor (air) Q factor (buffer)

SU-8 (Au coated) 500 100 5.5 6.641 6.178 $16 $6TOPAS (Au pads) 500 100 4.5 4.774 4.810 $10 $4

710 F.G. Bosco et al./ Microelectronic Engineering 87 (2010) 708711

8/2/2019 bosco - 2010 - mee

4/4

[4] E.-T. Hwu, S.-K. Hung, C.-W. Yang, K.-Y. Huang, I.-S. Hwang, Appl. Phys. Lett. 91(2007) 221908.

[5] M. Nordstrm, S. Keller, M. Lillemose, A. Johansson, S. Dohn, D. Haefliger, G.Blagoi, M. Havsteen-Jakobsen, A. Boisen, Sensors 8 (2008) 15951612.

[6] A. Greve, S. Keller, S. Dohn, A. Kristensen, M. Cerruti, A. Majumdar, A. Boisen,Effects of Surface Structured Polymer Cantilever Fabricated by Nanoimprint

Lithography, International Workshop on Nonmechanical Cantilever Sensors,2009.

[7] M. Calleja, M. Nordstrm, M. Alvarez, J. Tamayo, L.M. Lechuga, A. Boisen,Ultramicroscopy 105 (2005) 215222.

F.G. Bosco et al./ Microelectronic Engineering 87 (2010) 708711 711