of 8
1
Supporting Information:
Holey
Graphene as a Weed Barrier for Molecules
Matt Gethers,
†,‡
John C. Thomas,
§
Shan Jiang,
§
Nathan O. Weiss,
Xiangfang Duan,
§*
William A. Goddard, III,
†,�,
,#*
and Paul S. Weiss
§,
*
Materials and Process Simulation Center, California Institute of Technology, Pasadena, CA
91125, United States.
Department of Biological Engineering, California Institute of Technology, Pasadena, CA,
91125, United States.
§
Department of Chemistry and Biochemistry and California NanoSystems Institute, University of
California, Los Angeles, Los Angeles, CA 90095, United States.
Department of Materials Science and Engineering, University of California, Los Angeles, Los
Angeles, CA 90095, United States.
Department of Applied Physics, Chemistry, California Institute of Technology, Pasadena, CA,
91125, United States.
Department of Chemistry, California Institute of Technology, Pasadena, CA, 91125, USA.
#
Department of Materials Science,
Chemistry, California Institute of Technology, Pasadena, CA,
91125, United States.
*psw@cnsi.ucla.edu (P.S.W.), wag@wag.caltech.edu (W.A.G.), xduan@chem.ucla.edu (X.D.)
2
Figure
S1
. (A) Original transmission electron microscopy image from Figure 2 before
segmentation. (B) An image histogram of the data in A
showing the intensity threshold cut off
used to create an image binary. (C) Resulting binary mask,
where graphene holes are separated
from the graphene layer. (D) Small
outlier artifacts in the image binary are removed. (E) The
diameters of the remaining holes are displayed in a bar graph, binned by diameter (10 Å
bin
width); we measure an average 37 ±
8 Å
hole size.
3
Figure S2
.
(A,B)
Scanning tunneling micrographs (I
tunneling
= 3 pA, V
sample
= -
1.0 V) of “holey”
graphene on Au{111}/mica directly after deposition from solution of water and acetone. Images
show protrusions and depressions, displayed as brighter and dimmer, respectively
. We attribute
the higher protrusions as solvent that has not desorbed from the holes, and depressions as holes
(without solvent) within the graphene overlayer. (C) After annealing at 100
̊C for 24
h, all
solvent is evaporated and only the depressions (holes) remain.
4
Figure S3.
(A) Scanning tunneling micrograph (I
tunneling
= 3 pA, V
sample
= -
1.0 V) of “holey”
graphene on Au{111}/mica with (B) a corresponding apparent height histogram. Masking
techniques, performed in MATLAB, enable “holey” regions and graphene regions to be isolated
and analyzed independently. (C) The image in A is segmented by apparent height. The graphene
layer is 2.1 ±
1.2 Å higher in average apparent height compared to
(D) the exposed Au region.
5
Figure S
4.
Scanning tunneling micrographs (I
tunneling
= 3 pA, V
sample
= -
1.0 V) of “holey” graphene
filled with 1-
adamantanethiolate (1AD) on Au{111}/mica, where the spacing between adjacent
1AD molecules and graphene atoms is recorded. Images of the molecules were first smoothed
and then analyzed using the Regionprops function in Matlab in the molecular regions
highlighted. The inserted molecular layer shows an average spacing (across multiple images) of
7.2 ±
1.1 Å, while the graphene mask shows an average spacing of 5.0 ±
1.1 Å.
6
Figure S5
. Molecule Fitting Methodology. To determine nearest-
neighbor spacings between
molecules post-
1AD deposition, molecules were fit using the Regionprops function in Matlab. A
median filter is applied to remove intensity spikes, and then the region of interest is cropped for
analysis. The contrast of the cropped image is enhanced, and then the image is thresholded using
the Otsu cutoff. The cutoff was increased until sufficient segmentation was achieved. The
average adjustment was 0.16 where images were set to a grayscale. Finally, the center of each
segmented molecule was determined. The locations of these centers
were used to calculate
nearest-
neighbor distances. Fittings were also performed on regions that were analyzed
in
Fourier space to crosscheck results.
7
Figure S
6.
(A) Scanning tunneling micrograph (I
tunneling
= 3 pA, V
sample
= -
1.0 V) of “holey”
graphene filled with 1-
adamantanethiolate on Au{111}/mica with
(B) a corresponding apparent
height histogram. Masking techniques, performed in MATLAB, enable filled regions and bare
graphene regions to be isolated and analyzed independently.
(C,D) The image in A is segmented
by apparent height and displayed. A 1-
admantanethiolate patch appears on average 1.1 ±
0.5 Å
than the graphene layer.
8
Figure S
7.
(A, B) Scanning tunneling micrographs (I
tunneling
= 3 pA, V
sample
= -
1.0 V) of “holey”
graphene on Au{111}/mica after a second 1-
adamantanethiolate vapor
deposition for 24 h.
Each
sample was regenerated, prior to the second deposition step, by
annealing at 250
̊
C. Images
depict 1AD molecules within a “holey” graphene framework.