Towards airborne nanoparticle mass spectrometry with nanomechanical string resonators

Towards airborne nanoparticle mass spectrometry withnanomechanical string resonators

Silvan Schmid, Maksymilian Kurek and Anja Boisen

Department of Micro and Nanotechnology, Technical University of Denmark, DTU Nanotech,Building 345 East, DK-2800 Kgs. Lyngby, Denmark

ABSTRACT

Airborne nanoparticles can cause severe harm when inhaled. Therefore, small and cheap portable airbornenanoparticle monitors are highly demanded by authorities and the nanoparticle producing industry. We proposeto use nanomechanical resonators to build the next generation cheap and portable airborne nanoparticle sensors.Recently, nanomechanical mass spectrometry was established. One of the biggest challenges of nanomechanicalsensors is the low efficiency of diffusion-based sampling. We developed an inertial-based sampling method thatenables the efficient sampling of airborne nanoparticles on a nanomechanical sensor operating directly in air. Wemeasured a sampling rate of over 1000 particles per second, for 28 nm silica nanoparticles with a concentration of380000 #/cm3, collected on a 500 nm wide nanomechanical string resonator. We show that it is possible to reacha saturated sampling regime in which 100% of all nanoparticles are captured that are flowing in the projectionof the nanostring. We further show that it is possible to detect single airborne nanoparticles by detecting 50 nmAu particles with a 250 nm wide string resonator. Our resonators are currently operating in the first bendingmode. Mass spectrometry of airborne nanoparticles requires the simultaneous operation in the first and secondmode, which can be implemented in the transduction scheme of the resonator. The presented results lay thecornerstone for the realization of a portable airborne nanoparticle mass spectrometer.

Keywords: airborne nanoparticles, mass spectrometry, gravimetry, nanomechanical resonators, NEMS, inertialsampling

1. INTRODUCTION

The industrial use of engineered nanoparticles has increased dramatically in recent years, raising the risk of humanexposure to nanomaterials.1 When inhaled, engineered airborne nanoparticles can cause severe harm such aspulmonary inflammation, asthma, chronic obstructive pulmonary disease, and lung cancer when inhaled.2 Low-cost, portable personal airborne nanoparticle monitors have been demanded for years by authorities and thenanoparticle producing industry.3

Recently, nanomechanical mass spectrometry was established by Hanay et al.4 A nanomechanical resonatoris oscillating in ultrahigh vacuum and it is bombarded with the species to be analyzed. Analyte molecules ornanoparticles hitting the resonator increase its mass which results in a decrease in the resonance frequency. Fromthe gravimetric frequency changes, the mass of individual adsorbates can be calculated4–7 and from many suchmeasurements a mass spectrum can be derived.

In this work we present a nanomechanical mass sensor for the detection of single airborne nanoparticles,similar to the nanomechanical mass spectrometer of Hanay et al. But instead of operating in vacuum, whichrequires heavy and expensive vacuum equipment, our sensor operates at atmospheric pressure. This facilitatesthe design of a cheap and portable airborne nanoparticle mass spectrometer.

One of the biggest challenges with nanomechanical mass sensors is the small probability that a sampleentity will hit the small resonator surface purely by diffusion.8 To collect airborne species, the generally poorsampling efficiency has been improved by dielectrophoretic attraction,9 electrodynamic precipitation10 or by

Further author information: (Send correspondence to S.S.)S.S.: E-mail: [email protected]

Invited Paper

Micro- and Nanotechnology Sensors, Systems, and Applications V, edited by Thomas George, M. Saif Islam, Achyut K. Dutta, Proc. of SPIE Vol. 8725, 872525 · © 2013 SPIE

CCC code: 0277-786X/13/$18 · doi: 10.1117/12.2013682

Proc. of SPIE Vol. 8725 872525-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 06/11/2013 Terms of Use: http://spiedl.org/terms

a! b!

Diffusion!

Inertial Impaction!

ED∝ u -2/3 !

EI∝ u !

Figure 1. (a) Schematic drawing of the nanomechanical resonantor for airborne nanoparticle detection. The nanoparticlesare streaming through an orifice in the sensor chip and pass the nanostring resonator at velocity u. (b) The suspendednanoparticles are collected on the string either by Brownian diffusion or inertial impaction, with a single filter-fibercollection efficiency ED and EI , respectively. The collected nanoparticles are detected gravimetrically.

partial vacuum.11 Nevertheless, sampling times of several minutes to hours are typically required in order tocollect a small number of particles.

We have recently introduced an inertial-based sampling method for the efficient collection of airborne nanopar-ticles onto a nanomechanical mass sensor.12 With this new method it is possible to overcome the low analytecapture probability of diffusion-based sampling. The collection efficiency is strongly particle size dependent. Herewe present a single filter-fiber collection efficiency of 100%. The operation in this saturated regime is importantto perform quantitative mass spectrometry of heterogeneous airborne nanoparticles of different size. This effi-cient inertial sampling method sets the cornerstone for the realization of nanomechanical airborne nanoparticlemass spectroscopy.

A schematic of the presented sensor is depicted in Fig. 1a. The sampling method is based on collectionmechanisms known from air filtration. The aerosol stream is flowing through an orifice in the sensor chipthereby passing the nanomechanical resonator, also called nanostring. The airborne nanoparticles collect onthe nanostring by diffusion and/or inertial impaction, depending on their momentum (Fig. 1b). The nanostringbasically acts as a single filter-fiber. At low velocities, typically applied in air filtration and sampling (< 10 m/s),diffusion-based collection of airborne nanoparticles is predominant. This inefficient collection regime can beovercome by increasing the nanoparticle momentum, that is by increasing the air velocity through the orifice. Inthe inertial sampling regime the nanoparticles gained enough momentum to leave the air streamlines and impingeon the nanostring instead of flowing around it. The collected nanoparticles are detected gravimetrically, thatis by measuring the sudden frequency shifts of the nanostring’s resonance frequency induced by a nanoparticlecollection event.

2. THEORY

All models of air filters are based on calculations and assumptions made for single filter-fibers. The first reliablemodels describing the collection efficiency of single filter fibers were developed in the 50’s among others mostfamously by Irving Langmuir. Most models are only valid for a limited range of boundary conditions, in whichthey were empirically obtained by fitting with experimental data. This makes it difficult to use them for discussingquantitative experimental data. But there are some fundamental physical relationships in all models which allowa qualitative discussion. In this section we describe accepted models for diffusive and inertial precipitation onfilter-fibers.13 We are going to use these models as references to discuss the measured data with the resonantsingle filter-fibers.



The single filter-fiber collection efficiency is defined as the fraction of particles deposited on the filter-fiber tothe total number of particles that would have passed through the fiber if they had moved on straight lines. Asolution for the single filter-fiber collection efficiency due to Brownian diffusion was first developed by Langmuirin 194214–16

ED = βPe−2/3 = β

(dfu

D

)−2/3

(1)

where df is the fiber diameter, u the air velocity, β is an model dependent factor, and D is the particle diffusioncoefficient

D =kBTCc

3πνdp(2)

with the Boltzmann constant kB and temperature T . The value for β approximately calculated by Langmuir waslater improved by a more precise model by Natanson in 1957.16,17 Since then more models have been developed,most of them taking into account the flow field disturbance due to the dense proximity of fibers in an air filter.All models describe the collection efficiency based on the Peclet number Pe.15,18,19 It has been shown thatNatanson’s model is accurate for low Reynolds numbers16,18 with

β = 2.9(2− lnRef )−1/3. (3)

The models for single filter-fiber collection efficiency by inertial impaction are of empirical nature. All ofthem are based on the Stokes number

Stk =ρpd

2pCcu

18νdf(4)

where ρp is the particle mass density, dp the particle diameter, df the fiber diameter, u the air velocity, ν theair viscosity, and Cc the Cunningham factor correcting Stokes’ law for small particles (dp < λ, mean free pathof air molecules)

Cc = 1 +λ

dp[2.34 + 1.05 exp (−0.39

λ

dp)] (5)

The collection efficiency can be described by13,20

EI =2(Stk)J

Ku2(6)

where J is an empirical dimensionless term

J = 29.6

(dpdf

)2

− 27.5

(dpdf

)2.8

(7)

which is valid if dp/df < 0.4. Ku is the dimensionless Kuwabara hydrodynamic factor correcting for the flowfield distortion caused by neighboring fibers. The factor is based on the fiber volume fraction α in the filter

Ku = − lnα2− 0.75 + α− α2

4. (8)

We define the total collection efficiency Ec as the sum of the collection efficiencies due to Brownian diffusionED and inertial impaction EI . For a qualitative discussion, the collection efficiencies are simplified to

Ec = ED + EI = a1(dfu)−2/3 + a2u/df . (9)

A multitude of analytical and often empirical expressions for the collection efficiency have been developedover the years. Equation (9) represents the common denominator of the majority of models. Even though themodels for collection efficiency are based on cylindrical filter-fibers, equation (9) is also true for rectangular



i10 μm"

Figure 2. SEM image of a 500 nm wide SiN nanostring.

fibers.21 It can be seen that the collection efficiency is maximum for narrow fibers, which is supporting the useof nanomechanical filter-fibers. The number of particles collected per time is given by

Nt = dfLCpEcu = LCp

(a1(dfu)1/3 + a2u

2)

(10)

with the particle number concentration Cp, and the fiber length L. It can be seen that there are two samplingregimes. At low velocities diffusion is dominant with (u1/3) and at higher velocities most particles are collectedby inertial impaction with (u2). Thus the number of collected particles can be maximized by increasing thevelocity of the nanoparticles until the momentum becomes large enough such that the nanoparticles escape theair streamline and collect on the fiber by inertial impaction. In order to increase the aerosol velocity the chipsfeature a small orifice through which the air is accelerated. The nanomechanical resonant filter-fiber is spannedover the orifice and the aerosol passes the fiber perpendicularly.

Assuming that the fiber surface is homogeneously covered with nanoparticles it can be modelled as a lumpedelement linear resonator with a resonance frequency fr. For low damping and small mass changes, the masssensitivity can be written as

df

dm= − fr

2m0(11)

where m0 is the mass of the fiber. For monodisperse nanoparticles with an individual mass ∆m, the change inmass is equal to

dm = Nt∆m dt. (12)

Based on (11), (12), and (10), and assuming that the particle collection is dominated by inertial impaction, thegravimetric frequency change per time then becomes a linear function of the particle concentration Cp(t)

df

dt= −1

2

∆m

m0frNt = −1

2

∆m

m0frLCpdfEIu. (13)

3. EXPERIMENTAL

3.1 Fabrication of nanomechanical resonators

The nanomechanical resonators were fabricated by standard cleanroom micro processing, similar to the processdescribed in.22 100 nm silicon nitride (SiN) was grown on double sided polished silicon wafers by low pressurechemical vapor deposition (LPCVD). The silicon nitride layers were patterned by deep-UV projection lithographyand structured with a deep reactive ion etcher. The front side with the resonator pattern was covered with a500 nm thick layer of plasma enhanced chemical vapor deposited (PECVD) silicon nitride. The orifice was thenetched through the wafer from the backside with an anisotropic KOH etch. The front side PECVD was removedin buffered HF, releasing the resonators. Finally, a 50 nm thick Al layer was evaporated on the front-side. ASEM image of an uncoated nanomechanical SiN resonator is shown in Fig. 2



Laser-Doppler vibrometer

Aerosol In

Aerosol Out

Sensor chip

Magnet

Aerosol In

Aerosol Out

Actuation

Magnets Sensor chip

Filter

a! b!

Figure 3. a) Schematic drawing of aerosol measurement chamber. b) Photograph image of an aerosol measurementchamber.

3.2 Measurement setup

The sensor chip was mounted in an aluminium packaging providing electrical connection, a window for opticalreadout and connections allowing the aerosol flow through the chip’s orifice. The mechanical resonators weredriven magnetically by placing them perpendicularly to a static magnetic field and passing an alternating current.The magnetic field was created by neodymium magnets. The vibration of the nanomechanical resonator wasread-out by a laser Doppler vibrometer (MSA-500 from Polytec GmbH), as schematically depicted in Fig. 3a. Aphotograph of the aluminium measurement chamber is shown in Fig. 3b.

The resonator was driven as an oscillator with a phase-locked loop (PLL) with a digital lock-in amplifier(HF2LI from Zurich Instruments with PLL option). The resonance frequency was detected with a frequencycounter (53131A from Agilent Technologies) and recorded with a Labview program. The aerosol was producedwith an electrospray aerosol generator (EAG 3480 from TSI), neutralizing the aerosol with a radioactive ionizer(Polonium source). The air for the aerosol was filtered with a filtered air supply (3074B from TSI). The aerosol wasgenerated from 28.0± 3.2 nm23 silica particles (LUDOX TM-50 colloidal silica from Sigma-Aldrich), and 50 nmAu particles (Sigma Aldrich). The nanoparticles were mixed in 20 mM ammonium acetate buffer with pH 8 forthe use with the electrospray. The aerosol concentration was measured with a diffusion size classifier (DiSCminifrom Matter Aerosol) with an accuracy of ±30%, a maximal number concentration of 1×106 particles/cm3, anda time resolution of 1 s. The flow rate through the chip was measured with a mass flow meter (4140 from TSI)with an accuracy of ±0.005 L/min. The entire measurement setup is depicted in Fig. 4.

4. RESULTS & DISCUSSION

Fig. 5a shows the resonance frequency response of a nanostring with and without an aerosol flow through thechip orifice. As soon as the flow starts the nanoparticles precipitate on the string. This steady addition of massresults in a linear frequency slope, as described by (13). Fig. 5b shows the frequency slopes of the nanostring forvarying velocities. At low velocities (u < 5 m/s), the nanoparticles are collected on the nanostring by diffusion(uEc ∝ u1/3). At higher velocities (u > 5 m/s) the nanoparticles gain momentum and inertial impaction becomesdominant (uEc ∝ u2). At u = 50 m/s, the collection efficiency saturates (Ec = 1), and the number of collectedparticles per time becomes a linear function of u (uEc ∝ u). At u > 80 m/s the nanoparticles are blasted off orbounce off the string and the collection efficiency degenerates. From (13) the number of nanoparticles collectedper second Nt can be calculated. The numbers are plotted on the right y-axis of Fig. 5b. At high velocities, up to1000 particles per second are collected. This is orders of magnitudes more than current state-of-the-art airbornenanoparticle sampling techniques on micro or nanomechanical sensors with collection events in the minute cyclerange.

From this experiment it can be seen that it is possible to efficiently collect nanoparticles by inertial impactionby increasing the aerosol velocity. The efficiency by which nanoparticles are collected is a function of theirmomentum and size. Therefore for a sensor application it is important to operate in the saturation regime where



Lock-in amplifier!

Laser-Doppler!vibrometer!

Out!

Out!

In!

Measurement !chamber!

Electrospray!aerosol!

generator!

Air filter!

CO2!

Air!

Excess!aerosol!

Aerosol!Mixing !volume!

Vacuum!

Mass flow !meter!

Air filter!

Reference!sensor!

Flow!controller!

Frequency!counter!

Attenuator!

Figure 4. Schematic overview of experimental setup.

a! b!

28 nm silica nanoparticles!

Figure 5. Frequency response of a 500 nm wide, 100 nm thick, and 100 m long silicon nitride nanostring coated with50 nm of aluminium. (a) Frequency response to a sudden stream of a 28 nm silica aerosol flowing through the chip. (b)Frequency shift as a function of aerosol velocity.

nanoparticles are captured with a single fiber collection efficiency of 100%. From the slope of the linear fit ofthe saturated regime a particle concentration of Cp = 390000± 130000 #/cm3 can be calculated based on (13).This corresponds well with Cp = 380000± 110000 #/cm3 measured with the commercial reference sensor.

For recording a mass spectrum of airborne nanoparticles, it is crucial to be able to measure the mass of singlenanoparticles. In Fig. 6 the frequency response of a nanostring exposed to 50 nm Au nanoparticles is shown.The frequency jumps for the impact of single nanoparticle is clearly visible. The frequency shift of a string withtotal mass m0 due to a single mass ∆m is given by7

∆f = −fr∆m

m0sin2

(nπxL

)(14)

where fr is the eigenfrequency of the nth mode, L is the string length and x is the landing position of theparticle along the string length. The maximum frequency shift of a string due to a single mass is then given by∆f = −fr∆m/m0, which results in ∆f = −3400± 950 Hz. This corresponds well with the measured maximumfrequency jumps seen in Fig. 6. As is obvious from (14), ∆f is a function of the landing position of each individual



50 nm Au nanoparticles!

Figure 6. Frequency response of a 250 nm wide, 100 nm thick and 20 µm long nanostring exposed to 50 nm Au nanopar-ticles. The frequency jumps due to single nanoparticle collection events are clearly visible.

particle on the string. For the recording of the mass spectrum, ∆m can be evaluated by measuring the frequencyjump of the first and second mode.4–7

5. CONCLUSION

We have shown that airborne nanoparticles can be efficiently collected on a nanomechanical resonator by inertialsampling. The collection is most efficient at high aerosol velocities when nanoparticles gain sufficient momentumto leave the streamline and impinge on the resonator. It is possible to collect 28 nm silica nanoparticles witha single filter-fiber collection efficiency of up to 100%. In this saturated collection regime, up to 1000 airbornenanoparticles are collected per second from an aerosol with a concentration of 380000 #/cm3. State-of-the-art micro and nanomechanical nanoparticle sensors collect a few particles per minute. Thus, inertial samplingfacilitates the development of efficient sensors.

The silicon nitride nanostrings are sensitive enough to measure sudden frequency shifts due to the collectionof individual nanoparticles. For a future application in order to obtain a nanoparticle mass spectrum, themass of individual particles can be determined in real-time by applying a dual-mode operation of the resonator.Therefore the frequency shift of the first and the second bending mode is measured for every particle collectionevent. By downscaling the resonator it is possible to detect smaller single nanoparticles. The nanomechanicalsensor presents a feasible concept for low-cost portable monitors for the detection and characterization of airbornenanoparticles.

Acknowledgements

The research leading to these results has received funding from the European Community’s Seventh FrameworkProgramme (FP7/2007-2013) under grant agreement nr. 211464-2 and from the Villum Kann Rasmussen Centreof Excellence NAMEC under Contract No. 65286. Thank also goes to Christof Asbach and Luis GuillermoVillanueva for the valuable discussions.

REFERENCES

[1] Oberdorster, G., “Safety assessment for nanotechnology and nanomedicine: concepts of nanotoxicology.,”Journal of internal medicine 267, 89–105 (Jan. 2010).

[2] Li, J. J., Muralikrishnan, S., Ng, C.-T., Yung, L.-Y. L., and Bay, B.-H., “Nanoparticle-induced pulmonarytoxicity.,” Experimental biology and medicine 235, 1025–1033 (Sept. 2010).

[3] Maynard, A. D., Aitken, R. J., Butz, T., Colvin, V., Donaldson, K., Oberdorster, G., Philbert, M. A.,Ryan, J., Seaton, A., Stone, V., Tinkle, S. S., Tran, L., Walker, N. J., and Warheit, D. B., “Safe handlingof nanotechnology,” Nature 444(16), 267–269 (2006).



[4] Hanay, M. S., Kelber, S., Naik, A. K., Chi, D., Hentz, S., Bullard, E. C., Colinet, E., Duraffourg, L., andRoukes, M. L., “Single-protein nanomechanical mass spectrometry in real time,” Nature Nanotechnology 7,602–608 (Aug. 2012).

[5] Dohn, S., Svendsen, W., Boisen, A., and Hansen, O., “Mass and position determination of attached particleson cantilever based mass sensors,” Review of Scientific Instruments 78(10), 103303 (2007).

[6] Dohn, S., Schmid, S., Amiot, F., and Boisen, A., “Position and mass determination of multiple particlesusing cantilever based mass sensors,” Applied Physics Letters 97(4), 044103 (2010).

[7] Schmid, S., Dohn, S., and Boisen, A., “Real-time particle mass spectrometry based on resonant microstrings,” Sensors 10(9), 8092–8100 (2010).

[8] Squires, T. M., Messinger, R. J., and Manalis, S. R., “Making it stick: convection, reaction and diffusion insurface-based biosensors.,” Nature biotechnology 26, 417–26 (Apr. 2008).

[9] Wasisto, H. S., Merzsch, S., Waag, A., Uhde, E., Salthammer, T., and Peiner, E., “Airborne engineerednanoparticle mass sensor based on a silicon resonant cantilever,” Sensors and Actuators B: Chemical 180,77–89 (Apr. 2013).

[10] Lin, E.-C., Fang, J., Park, S.-C., Johnson, F. W., and Jacobs, H. O., “Effective localized collection andidentification of airborne species through electrodynamic precipitation and SERS-based detection,” NatureCommunications 4, 1636 (Mar. 2013).

[11] Hajjam, A., Wilson, J. C., Rahafrooz, A., and Pourkamali, S., “Fabrication and characterization of ther-mally actuated micromechanical resonators for airborne particle mass sensing: II. Device fabrication andcharacterization,” Journal of Micromechanics and Microengineering 20, 125019 (Dec. 2010).

[12] Schmid, S., Kurek, M., Adolphsen, J. Q., and Boisen, A., “Real-time single airborne nanoparticle detectionwith nanomechanical resonant filter-fiber,” Scientific Reports 3, 1288 (Feb. 2013).

[13] Hinds, W. C., [Aerosol Technology ], Wiley-Interscience, New York, NY (1999).

[14] Langmuir, I., Rodebush, W. H., and Lamer, V. K., “Filtration of aerosols and the development of filtermaterials,” tech. rep., Office of Scientific Research and Development, Report. No. 865 (1942).

[15] Lee, K. W. and Liu, B. Y. H., “Theoretical Study of Aerosol Filtration by Fibrous Filters,” Aerosol Scienceand Technology 1, 147–161 (1982).

[16] Fuchs, N. A., [The mechanics of aerosols ], Pergamon Press (1964).

[17] Natanson, G., “No Title,” Proceedings of the USSR Academy of Sciences 112(100) (1957).

[18] Ingham, D. B., “The diffusional deposition of fibrous filters,” Langmuir 12(4), 357–365 (1981).

[19] Fuchs, N. A., Kirsch, A. A., and Stechkina, I. B., “A Contribution to the Theory,” Faraday Symp. Chem.Soc. 7 (1973).

[20] Yeh, H.-C. and Liu, B. Y. H., “Aerosol filtration by fibrous filters - I. Theoretical,” Aerosol Science 5,192–204 (1974).

[21] Fardi, B. and Liu, B. Y. H., “Efficiency of fibrous filters with rectangular fibers efficiency of fibrous filterswith rectangular fibers,” Aerosol Science and Technology 17(1), 45–58 (1992).

[22] Schmid, S., Jensen, K., Nielsen, K. K., and Boisen, A., “Damping mechanisms in high-Q micro and nanome-chanical string resonators,” Physical Review B 84, 1–6 (Oct. 2011).

[23] Orts-Gil, G., Natte, K., Drescher, D., Bresch, H., Mantion, A., Kneipp, J., and Osterle, W., “Characteri-sation of silica nanoparticles prior to in vitro studies: from primary particles to agglomerates,” Journal ofNanoparticle Research 13, 1593–1604 (Apr. 2010).



Towards airborne nanoparticle mass spectrometry with nanomechanical string resonators

Documents

Transcript of Towards airborne nanoparticle mass spectrometry with nanomechanical string resonators