Microfabricated tools for manipulation and analysis of magnetic microcarriers
Mark Tondra, Tony Popple, NVE Corp., Eden Prairie, Minnesota; USA Nikola Pekas, Rachel Millen, John Nordling, Marc Porter, Iowa State Univ.; Ames, Iowa; USA Albrecht Jander; Oregon State Univ.; Corvallis, Oregon; USA Magnetic Microcarriers 2004: July 20, 2004; Lyon, France
•Introduction / Magnetic Module Platform Motivation
•Giant Magnetoresistive (GMR) fabrication process
•Magnetic Carrier Detection Examples •Conclusions / Acknowledgements
Motivation for BioMagnetICs Platform
•The Military / Homeland Defense wants rugged, lightweight, cheap, rapid handheld bioassays. (Also: accurate, multi-functional, easy to use, etc.) •Eliminating light source and optical readout reduces the size and cost. •Detector is in assay chip, solid state, disposable •Magnetic field source is small, rugged, cheap •These features are attractive for consumer and clinical diagnostics •NVE will apply the same technology to medium and high density microarray applications.
Need Controllable and Uniform Field
•Both Detection and Manipulation tools demand the application of a known field •In experimentation and exploration phase, users need flexibility
•Magnetic excitation module is both a platform for experimentation and a basis for design of specialized device applications
Magnetic Biosensor MicroArray Magnetic Microcarrier as Label
1) Magnetic labels are linked to assay target
2) Target specifically binds to matching probe on sensor chip surface
Happlied
3) Applied magnetic field magnetizes labels, stray fields influence sensor
4) Magnetic Sensor measures concentration of beads on surface
Functionalized Surface Micro Chip Magnetic Sensor
Giant Magnetoresistance (GMR) Decouvert en France
Universite d’Orsay, Paris-Sud
Baibich, M.N. , Broto, J.M., Fert, A., Nguyen, F., Van Dau, Petroff, F.,Eitenne, P., Creuzet, G., Friederich, A., and Chazelas, J., “Giant Magnetoresistance of (001)Fe/(001)Cr Magnetic Superlattices,” Phys. Rev. Lett., vol. 61, 2472-2475 1988.
AF Pinning Layer
SpinValve
Magnetic (Co,Fe, Ni & alloys) Non-magnetic (Cu & alloys, Al2O3)
Multilayer 2 mm Sandwich
10 nm
20 mm
Idealized Resistance vs. Field
Voltage / Resistance
Resistance R0 + sinq
q
Sense Layer Pinned Layer
-2
-1
0
1
2
Field (Oe) A typical Giant Magnetoresistive (GMR) Structure is 0.01 mm thick x 2 mm wide x 100 mm long, thin metal film, sensitive in-plane.
Low Hysteresis Spin Valve (GMR type)
This is the type of response the GMR detector in your hard disk drive reader has
1.9 1.88 1.86
Volts
1.84 1.82 1.8 1.78
-100 I = 0.1 mA
-50
0 Applied Field (Oe)
50
100
Magnefluidics Process Cross Section
About 25 fabrication steps
PDMS Lid
Silicon Nitride Silicon Nitride
Fluid Channel Channel BCB 5 mm thick
Magnetic Bead
Channel BCB 5 mm thick
Contact well and bonding pad
etch stop
M1
Metal 1 1.5 mm Al
window nitride 200 nm
Metal 1 1.5 mm Al Giant Magnetoresistive (GMR) Films
window nitride
~ 500 nm Chemical-Mechanical Polished (CMP) Surface [or BCB + thin nitride]
Buried Metal (M0) 1.5 mm Al
Sputtered Silicon Nitride Silicon Nitride 200 nm Silicon Substrate
PDMS flow channel mold from SU-8
Up to 860 mm thick Ultra-high aspect ratio Wafer level processing Rapid prototyping
Other Process Features
Gold coating for thiol chemistry SiO, Si3N4, Al2O3, other attachment chem. Sensor may act as an electrophoretic tool Polystyrene or other polymer coating?
Wheatstone Bridge Design for Bioassay
Reference resistors for field and temperature compensation
2 Sense resistors that experience the sample Rsense1 or field being measured OutReference resistors do not experience I+
RRef2
Out+
sample
Vout = Out+ - Out-. If there is nothing to make the sense resistors behave differently than the refs, Vout = 0 Volts
RRef1
Rsense2
GND
Layout: Define Sensing Region
I+
100 mm x 100 mm sense region
Rsense1 OutRRef1
RRef2 Out+ Rsense2
GND
Layout: Place Sense Resistors
Place Rsense1
I+
Rsense1 OutRRef1
RRef2 Out+ Rsense2
GND
Layout: Place Sense Resistors
Rsense1
Add identical Rsense2
Rsense2
I+
The two sense resistors are interwoven to allow them to sense the same region and experience the same excitation field.
Rsense1 OutRRef1
RRef2 Out+ Rsense2
GND
Layout: Place Reference Resistors
I+
Rsense1
RRef2 Out+ Rsense2
Add two identical Rref
OutRRef1
GND
Layout: Add Al Interconnect
OutI+
I+
GND
Out+
Rsense1 OutRRef1
RRef2 Out+ Rsense2
Add interconnection metal (Aluminum)
GND
Layout: Add Al Interconnect
OutI+
I+
Rsense1 OutRRef1
RRef2 Out+ Rsense2
GND
Out+
GND
Define sensing region by selective etching and surface chemistry
Magnetic Bioassay Handheld Reader
• • •
•
•
Pocketsized Connects to Laptop AD card Accepts 24 pin “diving board” On-Board field = 100 Oe @ 250 mA Still a lot of room for miniaturization
Magnetic Excitation Module Magnetic Design
generates ~ 100 Oe / 250 mA
Sensor Ferromagnetic “core”
Coil
Die-Holding Printed Circuit Board
Design Basics “Diving 24 pin surface mount edge connector Narrow die area for fitting between in the excitation magnet gap
Board”
Wire bonds are potted such that sense pad is still exposed
BioMagnetIC System
•A to D card in laptop •Pocket-sized excitation module •Disposable sensor cartridges •Adaptable device development platform •R vs. H plots •Vout vs. time
Present Experiments: Ferrofluid Plug
Nikola Pekas, ISU Detecting ferrofluid plugs flowing over GMR sensors
Ferrofluid Plug Formation
Ferrofluid
PFD
50 µm
50 µm
Ferrofluid • Flow rate 1.0 µL/min • Flow rate 0.2 µL/min • Plugs formed at approx. 500 Hz • Plugs formed at approx. 50 Hz
Hx in Sensor Plane, Flowing Ferrofluid Plug
Simulation using “Amperes” magnetic modeling software
Effective GMR sensor area
Happlied = 15 Oersted
X direction,
|| GMR sense axis, || direction of flow Plug dimensions: 13 mm wide 18 mm deep 85 mm long
FerroTec 307 10nm Co particles ~1% by volume
GMR Sensor Architecture
•Two reference and two sensing GMRs configured as a Wheatstone bridge
E1
RR1
GND
RS1 RS2
Isrc
RR2
E2
Flow Detection Data
20040128_09195 --- data set 25, field +20 Oe, raw data 0.048 0.046 0.044 0.042 0.04 0.038 0.036 0.034 0 0.02 0.04 0.06 0.08 0.1 time (s)
GMR signal (V)
Bridge1 Bridge2 Bridge3
Surface Chemistry on Sensor Chip: Linearity
Rachel Millen, Toshi Kawaguchi,Marc Porter; Dept. of Chemistry Iowa State Univ.
[RMP/RBLANK] * 100%
1000 800 600 400 200
y = 96.9x +85.9 R2=0.9994
0
Surface Coverage of MP (MP/mm2)
2
4
6
8
10
With S/N = 3, the detection limit is 0.1 MP/µm2 Using 60 nm MP, 0.1 MP/µm2 corresponds to a detectable MP surface coverage of 0.1%
Magnetic Biosensor MicroArray High Sensitivity
(2.8 mm diam)
Courtesy L. Whitman, Naval Research Lab.
1) >3 logs of dynamic range 2) Better than 1 fMolar with fluidics 3) Single label detection is possible with NiFe Beads
Acknowledgments
•Colleagues at NVE: Kevin Jones, John Taylor, Loren , Cathy Nordman, Dexin Wang, Zhenghong Qian, Jim Daughton; At Iowa State Univ.: Toshi Kawaguchi, Heather Bullen, Mike Granger •Collaborators: Naval Research Lab (Lloyd Whitman, Cy Tamanaha, Mike Miller, et al.); Johns Hopkins Univ. (Dan Reich, Sasha Anguelouch); Lousiana State Univ. (Josef Hormez, Kun Lian, Zhenchun Peng, Franz Jost, Challa Kumar) •Funding: DARPA, National Science Foundation
Conclusions
•Magnetic detection with GMR works, demonstrated on dozens of magnetic bead types, single nano-wires, ferrofluids, and flowing plugs
•Prototype Magnetic Excitation Board, Sensors, and Software are Available
•Can be used for detection and sorting (See talk by Nikola Pekas) •Clear path for productization of low-density DNA and protein assays •We are eager to apply this technology to relevant problems www.nve.com/~markt/biomagnetics
Present Experiments: Ferrofluid DNA Label
John Fox, Lightools
-0.0322
-0.032
Series1 10 per. Mov. Avg. (Series1) 20 per. Mov. Avg. (Series1)
Sensor Voltage (Volts)
-0.0318
-0.0316
-0.0314
-0.0312
-0.031 1 30 59 88 117 146 175 204 233 262 291 320 349 378 407 436 465 494 523 552 581 610
AD counts (~0.1 secs. each)
Present Experiments: Nanowire Detection
Dan Reich, Sasha Anguelouch, et.al.; Johns Hopkins Univ. Making 150 nm diameter ferromagnetic nanowires Detection of a single nanowire with 100 micron “Sense Pad” Ultimate goal is to detect and manipulate cells
Courtesy Sasha Anguelouch, Johns Hopkins U.
Present Experiments: NRL
Shawn Mulvaney et. al. Doing surface chemistry under TSWG
Cy Tamanaha is making a Cell DNA detector under an NIH grant
Present Experiments: Ultra-Thin Dielectric
Shan Wang, G. Li, J. Kemp; Dept. of Mat. Science and Eng. & Electrical Eng. Stanford Univ.
Trying to detect very small single Co particles Has special scheme for ultra-thin GMR passivation Trying to functionalize 5 nm Co particles
Recent New Magnetic Assays
Quantum Design (San Diego) and Biological Defense Research Directorate have developed a single-analyte magnetically based assay for soldiers. Works like a pregnancy test, better quantification. Quantum Magnetics is using a SQUID magnetometer to do assays in microtiter plates.
Detection Summary
Can detect full range of magnetic particle size and property Superparamagnetic, paramagnetic, soft and hard ferromagnetic, needles, etc. Typical dynamic range is >3 logs Single particle detection is possible Measurement time < 1 second Detects both flowing and immobilized magnetic particles Early assay versions coming out
Microarray Format System Development
Low-Density Diagnostics
Medium Density Gene Expression
•About NVE
High Density DNA experiments
•Magnetic Sensing Motivation •Magnetic BioSensing Examples
•Microarray Format System Development
•Conclusions
Have Detector, Other Pieces are Needed
•Need a working sandwich-like assay •The sensor surface must be biochemically functionalized •The sample must be prepared and introduced to the sensor array
•A magnetic label must be functionalized for use in the assay, and be compatible with the fluid dynamics
Diagnostics are Nearest-Term Product
•Pack ~20 sensors onto a 1mm x 1mm disposable chip •5 x 4 array with 2 references, 200 mm spot diameter •Premium on easy sample handling •Use simple lateral flow formats like pregnancy test •Must add multiplexing and quantitativeness to compete
•Cost and functionality are main driving factors
Gene Expression Assay Makes Sense
•Array of 96 sensors (8 x 12, 100 mm spot diameter) will handle 30-gene assays nicely
•Goes to the limit of wires from the chip that are reasonable
•Waiting for Gene Expression experiments to define nice applications
GMR Digital Switch (AD005-02)
V+
Trim Pads
Ground
Vout
Can Magnetic Reader Replace a High Density Optical Scanner for Genomics?
•Cost of reader is drastically lower (~100x) •Other components are the same
•1536-well array with integrated circuitry is coming soon.
•Need commercial partner to go much further on higher density device.
Development Challenges
•Need to integrate magnetic sensor with sample introduction technology / fluidics •Need to adapt spot arraying technology; spotting, printing, electric field , or other. •Labels need to be within a few diameters of the sensors to be detected.
Full Lithographic Microfluidic Fab Process
•Current microfluidic manufacturing has too much manual work •Cost / size relationship demands smaller devices •Need integrated circuitry for high sensor count
NVE BioSensor Activity
•Began work on Bead ARray Counter (BARC) with Naval Research Lab (NRL) in 1997.
•Presently funded by DARPA and NSF to make magnetic biosensors - many collaborations
•NVE is the leader in developing / manufacturing this kind of device
NVE BioSensor Projects
Marc Porter PI, Microanalytical Instrumentation Center, Iowa State Univ. Detection of flowing magnetically labeled objects Detection of proteins and DNA on fixed chip surface 3 year project, up for renewal in Summer „03 DARPA BioMagnetICs Mark Tondra PI, Support ~ 5 groups also funded by same program Make sensor chips, generate circuit boards, wire bonding, detection system design and software Demonstrate magnetic sorting in flow system 0.5 years through a 3 + 2 year project NSF SBIR Phase IIB John Anderson PI (NVE) Apply “hard edge” GMR sensors to BioSensor product Beginning of 1 year project
NSF XYZ-on-a-Chip
Coordinates
Z
Happlied
Y
X
a = 0.5 mm v = 0.1 mm
1 mm
1 mm
Stray Fields from Magnetic Beads
Happlied
Sphere is uniformly magnetized, ideal dipole Bead Field ~ 1/r3 r
The stray field opposes the applied field in the sensor plane
GMR
Stray Fields: in-plane excitation
Happlied Need a sensor with bipolar output
The detected in-plane stray fields are in one
1.9 1.88 1.86
Volts
1.84 1.82
r
-100 I = 0.1 mA -50
1.8 1.78 0 Applied Field (Oe) 50 100
GMR
Vertical Excitation
Happlied
Resistance ()
2350 2300 2250 2200 2150 2100 2050 2000 1950 -400 -300 -200 -100 0 100 200 300 400
Applied Field (Oe)
Need a sensor with unipolar output
The detected in-plane stray fields are in many directions
GMR
Misalignment: vertical vs. in-plane
2350 2300 2250
1.9 1.88 1.86
Resistance ()
2200 2150 2100 2050 2000 1950 -400 -300 -200 -100 0 100 200 300 400
Volts
1.84 1.82 1.8 1.78
Applied Field (Oe)
-100 I = 0.1 mA
-50
0 Applied Field (Oe)
50
100
Hintended
“Error” Field Herror
Hintended Detected by sensor for vertical excitation
Herror
Not detected by sensor for in-plane excitation
Map of Stray Fields in the Sensor Plane
Hx(Bead) / Happ vs. X and Y
0.14 0.12 0.1 0.08 0.12-0.14 0.1-0.12 0.08-0.1 0.06-0.08 0.04-0.06 0.02-0.04 0-0.02 0.5 0.3 0.1 -0.1
HX(Bead) / Happ
0.06 0.04 0.02
Chi = Bead Radius = Vertical Sep. = Average Hbead = 0.05 0.5 0.1 0.0556
0
-0.5
-0.3
-0.1
0.1
-0.3
Y (mm)
0.5
X (mm)
0.3
-0.5
Output of SV sensor vs. bead concentration
GMR Response of 200br Sensor with Different %MP Drop Casts Normalized Baseline
110 105
Air Blank 0.001% MP 0.01% MP 0.1% MP
E (mV)
100 95 90 85 80 -100 -50 0 Field (Oe) 50 100
Courtesy R. Millen, M. Porter, Iowa State Univ.
Output of SV sensor vs. bead concentration
1.05
Response vs. Solution % MP
Slope of GMR Response from -24 Oe to 8 Oe
112 111 110
1.00
0.95 109 0.90 108 0.85
Slope Response at 8 Oe
107 106
0.80 0.00 0.02 0.04 0.06 0.08 %MP Drop Cast on Surface 0.10
Courtesy R. Millen, M. Porter, Iowa State Univ.
GMR Response at 8 Oe (mV)
NVE Corporation
Founded in 1989 as Nonvolatile Electronics Now traded on NASDAQ as NVEC About 60 employees 6000 ft2 clean room (class 100) Specialize in integrated magnetoresistive devices • FY ’03 revenues ~$9M, profitable
– Magnetic Sensors [biomedical, industrial] – Digital signal couplers – Contract and Govt. Research and Development
• • • • •
NVE Highlights
• Licensed by Honeywell in 1989 to produce MagnetoResistive Random Access Memory (MRAM) • Introduced world’s first Giant Magnetoresistive (GMR) product in 1996 • Signal isolator product introduced in 1999 • Research & Development Thrusts
– – – – – Biosensors MRAM (MagnetoResistive Random Access Memory) Low-Field Magnetometers Non-destructive Evaluation Electronics and systems
NVE Fabrication Facility
PE 2400
Comptech
Magnetic Bioassay Reader Block Diagram
Laptop / PC
Edge Connector
Connector Board
GMR Array Chip
64 x 1 Multiplexer
12 Bit A to D Converter
Diff. Amp.
Digitizer Board
Current Amp Coil Board