Enhancement of High-Performance Graphene Biosensors for

Cancer Detection
VictorAcero ,ShotaSando ,TianhongCui 1 2 2

DepartmentofMechanicalandNuclearEngineeringThePennsylvaniaStateUniversity 1

DepartmentofMechanicalEngineeringUniversityofMinnesotaTwinCities 2

Backgrounds Goals
Elevated alpha-fetoprotein (AFP) levels are tied to issues with the Develop graphene biosensor capable of detecting low levels of alpha-
liver cancer and ovarian cancer1. Enhanced detection of AFP would fetoprotein. We will use graphene because it is a 2D material, and
allow for quicker medical responses. therefore hyper-sensitive to its environment
Current tests such as ELISA are complicated, slow, and require fine Detect AFP concentrations in a matter of minutes
tuning for antigen being detected Lower 1/f noise levels via graphene-gold interface redesign
FabricationProcess

Start Pattern Gold onto Wafer Pattern Graphene onto sensing area Pattern KMPR window onto sensor

Photolithography is used to: Testing With and Without KMPR Fully fabricated
Pattern to basic gold electrode shape Graphene Bio-Sensor
Current
Pattern graphene onto sensing area Current 70% Success
Pattern protective KMPR coating
10 mm width
13 mm height
.55 mm depth
GrapheneLayerbyLayerSelfAssembly pHTesting
Graphene LbL Self-Assembly uses negative/positive polyelectrolyte and Graphene changes resistance
graphene solutions to form few-layer graphene depending on surroundings
Holes are main charge carriers in
PSS PDDA graphene.
2 Nearby negative charges decrease
cycles Resistance
Nearby Positive charges increase
resistance
Graphene

Graphen Polyelectrolytes
PDDA
e 5 Tunable graphene thickness
cycles Tunable optical, mechanical, and electrical properties
Graphene self assembly is possible on various substrates Graph 1 As pH is increased the resistance
Large surface area graphene increases. Sensitivity in the reverse in poor
Mass production is realistic unlike with traditional methods because of H+ contamination on surface.

AFPDetectionTest Functionalization Conclusions

1. Coat Graphene with Poly-L-lysine AFP Sensing: Our data showed
Graph 2: AFP concentrations Helps Anti-AFP bind to Graphene
the sensors that were fabricated
1 pg/ml
were slowly increased by one
5200
order of magnitude at a time.
2. Functionalize surface with Anti-AFP
had at least the same sensitivity as
10 pg/ml

Large spikes in graph are due 3. Fill gaps with Bovine Albumin Serum
100 pg/ml
5000
to pressure changes when Keeps unbound AFP and Ions from lodging between Anti-AFP ELISA under ideal conditions. Our
Resistance (Ohms)

1 ng/ml
10 ng/ml
changing solutions. 4. Introduce solution with AFP
4800 100 ng/ml
Increase concentration slowly sensors also responded in a matter
(Inset) : Normalized AFP of seconds to elevated AFP
4600
Sensing data demonstrates
concentrations, as opposed to
-0.03

-0.04

the relationship between

Normalized Resistance Shift

-0.05

AlphaFetoprotein
within hours, as with ELISA.
-0.06

-0.07

4400 -0.08

-0.09
resistance and AFP
concentration
-0.10

-0.11

-0.12

4200 1E-13 1E-12 1E-11 1E-10 1E-9 1E-8 1E-7 1E-6

AFP Concentration (g/ml)

BovineSerumAlbumin Noise Reduction: Reduction in

AntiAlpha
0 500 1000
Time (s)
1500 2000 2500 3000
Fetoprotein
GrapheneandPolyLlysine the overall surface area contact
between the gold-graphene
GoldGrapheneInterfaceRedesign NoiseReductionData interface showed a reduction in 1/f
Graph 3: Dynamic 1/f Noise No
NoStrip
Strip 31 Strips: 1/3 Original Surface Area
noise levels.
0.0055 31 Strip
31 Strip
signal analyzer was
0.0055
No Strip 95Strip
95 Strip
20 um strip width / 40 um gap
Acknowledgements
0.00500.0055
0.0050 31 Strip 950 Strip
950 Strip
used to determine the 95 Strip
Power Spedctrum Density (V^2/Hz)

95 Strips: Original Surface Area

0.0045
0.0050
Power Spedctrum Density (V^2/Hz)

0.0045 950 Strip

0.0040
PSD and find 1/f noise.
Power Spedctrum Density (V^2/Hz)

0.0045 1

0.00400.0040 0.0035
0.1

20 um strip width / 20 um gap NNIN/NSF

Power Spectrum Density (V^2/Hz)

0.01

However a DC voltage 0.00350.0035 0.0030

1E-3
1E-4
1E-5 950 Strips: Original Surface Funding and academic support throughout the
0.0025
source is used when 0.0030
0.0030

0.0025
0.0020
1E-6
1E-7
Area project
detecting AFP, so the
1E-8
0.0025 0.0015

2 um strip width / 2 um gap National Science Foundation - Grant No. ECCS-

1E-9
0.0020
1E-10

0.00200.0015 0.0010

higher frequency noise

1E-11

0.0015
0.0010
0.0005
1E-12
1E-13 0335765
spectrum (White Noise)
1E-14
0.0000
0.0010
0.0005

0.0000
-0.0005
1 10 100
Frequency (Hz)
1000
UMN MRSEC
is not relevant to finding 0.0005-0.0005
0 2 4
Frequency (Hz)
6 8 10
Academic advising and professional
1/f noise for the sensor.
0 2 4 6 8 10
0.0000 Frequency (Hz) development
-0.0005
0 2 4 6 8 10
TIAN Lab
Resources and additional financial support
Frequency (Hz)
1/fdominatessignaltonoiseratioinnanoelectronicsand
Tianhong Cui/Shota Sando
(Inset) : The resolution at low frequencies is poor due to not having a greatlyaffectssensitivity.Sensorswereredesignedtohave Principle Investigator and Graduate Student
low-frequency amplifier, however it is still capable of demonstrating that reducedsurfaceareacontact Mentor respectively, who provided guidance
the reduction in surface area was capable of reducing overall 1/f noise Contactresistancecontributegreatlyto1/f throughout the summer. They advised me on
in the sensor. Reductioninoverallsurfaceareareducedoverallnoise research and aided in professional development