University College Dublin

Final year project NanoBio

Author: Oisin Maguire Supervisor: Dr. J Rice

13th September 2013

Abstract In this experiment a thin layer of silver is plated onto a glass slide and is illuminated using a green laser of which causes surface plasmons to form across the thin silver surface and these surface plasmons causes a blinking eect across the surface of the metal of which are recorded using a charge coupled device CCD and a lter as to remove the light scattered o the metal surface through the Rayleigh scattering mechanism. In this experiment we vary the temperature of the silver surface as to determine if there is a temperature dependence on the Raman mechanism. As the end of Moores law is approaching the need for a dierent strategy to increase are technologically progressing society is increasing and as such many professionals in science and engineering suspect that the next revolution is going to be in the area of nano-bio and as such are understanding on the ner points on the nano-scale are being tested currently by multiply research groups around the world. This will ineable open up new ways we interface with technology and diagnosing people with varies disease will become easier and the treatment of what would be a terminal illness today will be a curable illness tomorrow.

Contents
I Introduction 3
3

1 Theory

II

Experiment

5
5 7

2 Experimental Set up 3 Experimental Procedure

III

Data

11
11 14

4 Data and Data Analysis 5 Code

IV

Discussion

19

Part I

Introduction
1 Theory
tering Stokes and anti-Stoke scattering in which the wavelength gets longer Stokes and one in which the wavelength gets shorter anti-Stokes as illustrated in the gure below. There are two types of processes to do with scattering Rayleigh scattering in which the wavelength of the incident light is equal to the wavelength of the emitted light and Raman scat-

Figure 1: Rayleigh and Raman mechanism

As the wavelength of light increases the energy of the photon decreases which is given by the formula below. c E=h (1)

of Raman scattered light to Rayleigh scattered light is very small of the order of 1014 . When there is no surface enhancement of the Raman scattering the signal to noise ratio becomes very signicant because they are of the same c where = , E is the energy of order of magnitude. the photon, is the wavelength of the photon, is the frequency of the To this end surface enhanced Raphoton, h is a constant of proportion- man spectroscopy SERs is the use of ality which is called Plancks constant a metal usually gold Au, silver Ag or and c is the speed of light in a vacuum. copper Cu as to form a standing elecThe Raman eect was discovered in tron wave surface plasmon in which 1928 by C. V. Raman and he gured the molecule that is being analysed out that the shift in the wavelength increases the amount of light emitted of the incident light depends on the that is Raman scattered. The molecule chemical structure of the molecules re- gains this ability by the increased elecsponsible for the scattering. The ratio tric and magnetic eld strengths on the 3

surface of the metal of which the molecule is residing this can increase the amount of light emitted by the Raman method by between 106 to as high as 109 which is a very large increase. The three types of metals used have slightly dierent uses and costs. Copper is the cheapest of the 3 metals but the intensity increase of the Raman mechanism is the lowest out of the three metals. Gold has the highest eciency of the three metals and is very resistant to air and other forms of corrosion however gold is the most expensive on the list of metals and as such silver was used as a good optimisation of resources. Over the past decade chemists have made periodic structures in the surface of the metal by pasting polystyrene spheres onto a thin gold foil. The polystyrene spheres line up in a hexagonal array. Then using electro chemistry they build up gold around the spheres of which then the spheres get dissolved away leaving behind a gold surface with voids that are the shape of spheres close to the surface. Theses spheres trap the incident light and help the absorption of the incident light, causing a greater surface plasmon density. This method alone can increase the Raman eect another thousand fold giving around a 1013 fold increase in the Raman mechanism than just the sample being analysed by it self. The optimal size of the spherical voids is of the order of a wavelength of the incident light. An easy way to think about what

a surface plasmon is, is to think about the reverse idea like a radio transmitter in which, say you have a large cylinder of metal and you apply an alternating voltage to the metal cylinder. This will cause a varying electric eld in time which acts like light and the frequency of the emitted radiation is the same as the frequency of the applied voltage to the metal cylinder. This is very much like the same as a surface plasmon but I be leave it is easier to grasp. Now for a surface plasmon light is incident on a piece of metal that results in a quantised oscillation of the electrons on the metal surface this is like having a voltage applied in the sense that it causes the metals to shift further from there expectation value for position. The diraction of any optical device from a telescope to a microscope is given by the following formula. d= 2nsin() (2)

Where d is the resolvable feature size, n is the refractive index of the medium and is half the angle from one side of the object to the other side as seen by the objective lens. nsin() is called the numerical aperture and is of order unity. The charged coupled device (CCD) in this experiment was cooled down to 20 degrees Celsius as at higher temperatures electron leakage due to thermal oscillation within the semiconductor would build up and causing a greater background error.

Part II

Experiment
2 Experimental Set up
oil is closer to that of glass than air and as such the spreading of the light due to the change in refractive index of the material is less signicant. An easy way to demonstrate that dierent wavelengths of light travel at dierent speeds in a material is to use white light and a prism as demonstrated below. The microscope used is an inverted oil submerged microscope as we are working very close to the diraction limit and as such the pocket of air introduced by a convex lens causes the light that is not coaxial to diverge away from the focal point of the lens and as such would causes a blurring of the image reducing image quality. Oil is used as the refractive index of

Figure 2: Prism demonstating dierent wavelengths of light travel at dierent speeds

The oil used with the microscope is to minimise the spreading of the different wavelengths of light i.e. in the gure above, the distance between the blue light and red light would be smaller as the area surrounding the prism would have a refractive index closer to that of the prism. Silver was plated on to a thin glass sheet using a vacuum chamber by va5

porising the silver using a large power supply as the power dissipated onto the silver is the volt amp across the silver. This silver vapour is then accelerated by using another power supply and the thin glass sheet is placed in between the anode and the cathode of this power supply as the silver will land and condense onto the glass sheet forming a layer of silver on the glass.

The glass sheets were cleaned with a clean cloth and then rinsed with acetone as acetone is a volatile liquid it vaporises quickly and leaves behind a clean glass sheet. A clamp was used as to attach the thin glass slide with the silver on the bottom side onto the bottom side of a temperature controlled piece of metal which adjusts the temperature of the sample and maintains an accurate temperature for the duration of the experiment. The temperature in the piece of metal is controlled electronically by

measuring the resistance in the metal and by varying the current, as the resistance of the metal will change as the temperature of the metal changes and knowing the temperature to resistance prole of the metal, the metals temperature can be measured accurately and the current through the metal can be altered to maintain a constant temperature. After the temperature of the metal reaches the desired temperature a waiting period of 5 minutes was used as to minimise small uctuations in the temperature of the sample.

Experimental Procedure
where the blink in the captured video is this is easily resolved by writing a program to plot the image of the frame of which the blink occurs. The gure below demonstrates the dierence between a blink present in an image and no blinks in an image.

As there is a nite bandwidth on the computer used, a section of the total size of the image was saved as this means more frames can be captured. This however will introduce a pixel shift between where the capturing software says a blink occurs and

Figure 3: Demonstration of a blink

The image on the left has no blink present however the image on the right has a blink present in the top left quadrant. To pin down the position of the pixel an educated guess was rst taken as the blink is about 0.35 of the entire image from the left and as there is 330 pixels across in this image the y-axis

slice was taken through the x = 116 axis rst and moved a few pixels to the left and right to nd the maximum value of the peak of the blink. The gure on the next page is of the y-axis slice of one of the silver samples at 20 degrees Celsius.

Figure 4: Y axis slice through silver at 20 degrees celcius

In the diagram above it is evident that the blink occurs at the y-axis coordinate of 100 therefore the blink occurs at the pixel in the .avi le at (x=116,y=100). All the data was recorded by a microscope with a kinetic cycle time of 35.804Hz that is to say we took images at a rate of one every 0.0279 seconds, the exposure time was 0.009 seconds for each image that is to say we essentially integrated over all the received photons over a period of 0.009 seconds. Throughout the exper-

iment the sample was heated to 40 degrees Celsius and cooled to 20 degrees Celsius several times and as this was happening three hot spots were monitored as to determine if the sample was drifting in anyway that is to say if the sample was transforming in a linear transformation or rotating the occurrence of these transformations could be monitored and discussed. Unfortunately there was no such rotations or transformation of any kind as the samples were heated and cooled.

Figure 5: Sanity check of hot spot locations

As a sanity check the gure above is the rst On the left side and last image On the right side taken of one of the samples to demonstrate that the hot spots stayed in there locality. The three black arrows in the gure above is to show the three hot spots I monitored throughout the experiment for this sample. As we are dealing so close to the diraction limit of the incident light onto the microscope the small changes in the shape of the lens causes the im-

age at one temperature to be slightly out of focus at a dierent temperature and as such during the experiment small changes were made in the focus of the image as to maintain the same aspect ratio of the image through the experiment. I did however try with one sample, not changing the focus of the image with a very drastic change in the image quality between a temperature at 20 degrees and 40 degrees as illustrated on the next page.

Figure 6: Focus dierence between 20 degrees and 40 degrees celcius

In the gure above the image on the left is a sample at 20 degrees Celsius and the image on the right is at 40 degrees Celsius this is to show the rapid change in image quality going between the two temperatures without

changing the focus on the microscope. This is the reason why I decided to change the focus of the microscope very slightly going between the dierent temperature.

10

Part III

Data
4 Data and Data Analysis
to identify the coordinates as described in the experimental procedures section. The gure below is a slice through a blink at 30 degrees Celsius. It is easier to nd blinks when you go through frames slowly and identifying the frame by the frame number and then taking slices through the image

Figure 7: Y axis slice through silver at 30 degrees celcius

A program was written as to determine the average value of the background of an image by taking an area with no blinks in this program a 10 by 10 matrix of pixels were added for each of the 800 frames and then divided by 80000 as to get an average value of the

background. To determine how the blinking changes with time a program was written as to graphically demonstrate the intensity of the blinking through time. The gure on the next page is of a silver sample at 20 degrees Celsius.

11

Figure 8: Blinking through time at 20 degrees celcius

The peaks in the above diagram are of the silver sample blinking, by taking the average background value for a .avi and counting only the blinks that occur above 1.5 time the background we can be sure to only add the peaks that ac-

tually correspond to the silver sample blinking. The gure on the next page is of the number of blinks recorded above the 1.5 times the background cut o point for each of the temperatures recorded.

12

Figure 9: Histogram of the number of blinks at varying temperatures

In the gure above it is clear that Celsius. This rapid change is discussed there is a very rapid change in the in the discussion section of the lab rebehaviour in the silver sample going port. from 20 degrees Celsius to 35 degrees

13

Code

Code to get a background value for the .avi close all ; clear all ; mov=a v i r e a d ( f 1 d . avi ) ; u=10; %u s e s 10 p i x e l s i n t h e x c o r d i n a t e p=10; %u s e s 10 p i x e l s i n t h e y c o r d i n a t e g =0; l a s t =800; %t h e r e i s o n l y 800 f r a m e s due t o %bandwidth l i m i t a t i o n f o r w=1: l a s t ; % f o r l o o p a c t i n g a s a sum o v e r f r a m e s f o r d=1:u ; % f o r l o o p a c t i n g a s a sum o v e r x c o r d f o r f =1:p ; % f o r l o o p a c t i n g a s a sum o v e r y c o r d [ im1 , map]= frame2im (mov(w ) ) ; %t a k e s a s t i l l o f %t h e . a v i g=g+d o u b l e ( im1 ( d , f ) ) / ( u p l a s t ) ; %sums i n t e n s i t y %a t p i x e l end end end background=round ( g ) %d i s p l a y s end r e s u l t s i n c e %im u s i n g u i n t 8 images I need a rounded %number a s i l l be s u b t r a c t i n g t h e %background from t h e o t h e r images

14

Code to plot a slice through a blink as well as give a still picture of the .avi at the blink time close all ; clear all ; mov=a v i r e a d ( f 1 d . avi ) ; a =100; %frame number o f b l i n k [ im1 , map]= frame2im (mov( a ) ) ; colormap ( map ) ; %t u r n s 3 matrix c o l o u r %i n t o f a l s e c o l o u r image ( [ 0 , 1 ] , [ 0 , 1 ] , im1 ) ; %r e s i z e s a x i s %o f image l a s t =300; %l a s t e n t r y i n t h e matrix %o f t h e width o f t h e image f o r i =1: l a s t x =10; %p o s i t i o n o f x a x i s s l i c e y ( i )=im1 ( i , x ) ; end figure ; plot (1: last , y , b ); h o l d on ;

15

Code to plot the time evolution of a blink close all ; clear all ; mov=a v i r e a d ( f 1 d . avi ) ; x =86; y =108; l a s t =800; V= [ 0 : 1 : l a s t ] ; f o r c =1: l a s t ; [ im1 , map]= frame2im (mov( c ) ) ; i f im1 > 0; V( c)=im1 ( x , y ) ; end p l o t ( c ,V( c ) , b ) h o l d on ; end x l a b e l ( Frame number ) ; ylabel ( Intensity ) ; t i t l e ( I n t e n s i t y o v e r time ) ;

16

Code to plot a histogram of the number of blinks above a cut o point close all ; clear all ; mov1=a v i r e a d ( 20 001. avi ) ; mov2=a v i r e a d ( 25 0001. avi ) ; mov3=a v i r e a d ( 30 0001. avi ) ; mov4=a v i r e a d ( 35 0001. avi ) ; mov5=a v i r e a d ( 40 001. avi ) ; l a s t =800; x1 =218; %The d i f f e r e n t c o o r d i n a t e s i s y1 =214; %b e c a u s e o f t h e change i n t h e x2 =142; %s i z e o f he matrix r e c o r d e d y2 =105; %by t h e m i c r o s c o p e x3 =130; y3 =92; x4 =120; y4 =78; x5 =145; y5 =150; V=0; V1=0; V2=0; V3=0; V4=0; V5=0; f o r c =1: l a s t ; [ im1 , map]= frame2im ( mov1 ( c ) ) ; [ im2 , map]= frame2im ( mov2 ( c ) ) ; [ im3 , map]= frame2im ( mov3 ( c ) ) ; [ im4 , map]= frame2im ( mov4 ( c ) ) ; [ im5 , map]= frame2im ( mov5 ( c ) ) ; i f im1 ( x1 , y1 ) > 70; V1=V1+1; end i f im2 ( x2 , y2 ) > 70; V2=V2+1; end i f im3 ( x3 , y3 ) > 70; V3=V3+1; end i f im4 ( x4 , y4 ) > 70; V4=V4+1; end i f im5 ( x5 , y5 ) > 70; V5=V5+1;

17

end end p l o t ( 2 0 , V1 , b , 2 5 , V2 , b , 3 0 , V3 , b , 3 5 , V4 , b , 4 0 , V5 , b ) x l a b e l ( Temperatures ) ; y l a b e l ( Number o f b l i n k s above 7 0 ) ; t i t l e ( On ) ;

18

Part IV

Discussion
In this section I will discuse the res- ber of blinks as the temperature of the ults of the experiment mainly the inter- sample is changed this is shown in the pretation of the histogram of the num- gure below.

Figure 10: Histogram of the number of blinks at varying temperatures

So why is there a sudden change in blinking as the temperature increases? Is it due to the orientation of the nanoclusters I can safely say no due to the fact that the material was heated and cooled several times with the same effect each time, is it just thermionic emission, also no as the temperature of the material is a good 1000 Kelvin o the black-body temperature to be in the near visible spectrum, also if it was thermionic the emission would be from the entire surface of the sample as oppose to just being from a small point on the surface. I suspect that the sudden change is due to either a

change in the size of the nano-cluster in eect absorbing more energy from the surface plasmons and thus discharging more often using the Raman mechanism or a quasi phase change or resonance between the incident light and the silver nano-clusters. An easy extension to further increase are understanding of this change would be to simply take more dierent temperature readings in the 20 to 30 degrees Celsius region but I suspect are knowledge of this process wouldnt increase suciently enough to justify the experiment. I would however try to isolate a nano-cluster of silver sus-

19

pend it in a vacuum chamber using an external electric-magnetic eld to stabilise it and using a laser as a pumping mechanism to increase the temperature of the nano-cluster the laser will also cause surface plasmons to form on the nano-cluster thus light will be emitted from the nano-cluster using the Raman mechanism. Analysing the light emitted by various sizes of nanoclusters and at various temperatures around the biological temperature the

silver nano-clusters may proved a step forward in understanding the properties of nano-clusters. On a side note a program was also written to take the intensity at a blink and add them if they are greater On time or less O time than a certain value which changes as we are using uint8 I started with 0 and went up to 255 the gure below is the On time On the left hand side and O times On the right hand side

Figure 11: On and O time

In gure 11 above the start and the end of the curve was removed as at the low values every point has a greater value than 0, and at the high end of the spectrum no events saturated the CCD and as such I though it was best to exempt them from the gure above. Figure 12 on the next page is with both

ends of the curve extending to 0 and 255 respectfully. In gure 12 it is clear why one would remove the start and end of the curves as the area of most interest has a linear Straight line relation on the log-log plot.

20

Figure 12: On and O times full

I excluded the Log-Log plots in the adds a dierent perspective but I be main parts of the lab report as I didnt leave that is all. The code for the Logsee how it would add to the report it Log plot is below. Code to plot a log-log plot of the .avi le at the blink site close all ; clear all ; clc ; mov=a v i r e a d ( 20 0 a . avi ) ; x =86; y =108; %y c o o r d l a s t =800; %l a s t frame u=10; %number o f p i x e l s i n x f o r e r r o r p=10; %number o f p i x e l s i n y f o r e r r o r g =0; %e r r o r count f o r w=1: l a s t ; f o r d=1:u ; f o r f =1:p ; [ im1 , map]= frame2im (mov(w ) ) ; g=g+d o u b l e ( im1 ( d , f ) ) / ( u p l a s t ) ; end end end background=round ( g ) ; %I need an INT a s im u s i n g UINT8 S l o p e =0; S t a r t =0; eNd=255; f o r b=S t a r t : eNd ; % ON q=0; V= [ 0 : 1 : l a s t ] ; f o r c =1: l a s t ; 21

[ im1 , map ] = frame2im (mov( c ) ) ; % GRAB A SINGLE FRAME V( c ) = im1 ( x , y) background ; i f V( c)> b ; q=q+1; end end h o l d on ; loglog (b , q , b ) ; h o l d on ; g r i d on ; xlabel ( Intensity ) ; y l a b e l ( Number o f counts ) ; t i t l e ( On times ) Q=p o l y f i t ( b , q , 1 ) ; S l o p e = S l o p e + Q( 1 , 1 ) ; end SON=S l o p e / ( eNd S t a r t ) ; SON figure ; SLope =0; s t a r t =0; enD=255; f o r b=s t a r t : enD ; %OFF q=0; V= [ 0 : 1 : l a s t ] ; f o r c =1: l a s t ; [ im1 , map ] = frame2im (mov( c ) ) ; % GRAB A SINGLE FRAME V( c ) = im1 ( x , y) background ; i f V( c)< b ; q=q+1; end end loglog (b , q , b ) ; h o l d on ; g r i d on ; xlabel ( Intensity ) ; y l a b e l ( Number o f counts ) ; t i t l e ( O f f times ) Q=p o l y f i t ( b , q , 1 ) ; SLope = SLope + Q( 1 , 1 ) ; end SOFF=SLope / ( enD s t a r t ) ; SOFF

22

Appendix References
[1] Chapter 18 Raman Spectroscopy revision 23, Anwar A. Bhuiyan [2] Prism Figure ,upload.wikimedia.org [3] Fluorescence blinking dynamics of silver nanoparticles and silver nanorod lms ,Xiangyang Wu and Edmin K L Yeow [4] Origins of Blinking in Single-Molecule Raman Spectroscopy ,Zhenjia Wang and Lewis J. Rothberg [5] Power-Law Blinking in the Fluorescence of Single Organic Molecules ,Jacob P. Hoogenboom Jordi Hernando Erik M. H. P. van Dijk Niek F. van Hulst and Maria F. Garcia-Parajo [6] Surface Plasmons and Surface Enhanced Raman Spectra of Aggregated and Alloyed Gold-Silver Nanoparticles ,Y. Fleger and M. Rosenbluh [7] Quanitative evaluation of blinking in surface enhanced resonance Raman scattering and uorescence by electromagnetic mechanism ,Tamitake Itoh Mitsuhiro Iga Hiroharu Tamaru Ken-ichi Yoshida Vasudevanpillai Biju and Mitsuru Ishikawa [8] Adaptive silver lms for surface-enhanced Raman spectroscopy of biomolecules ,Vladimir P. Drachev Mark D. Thoreson and Vladimir M. Shalaev [9] Silver nanoparticles and thin lms for strained-silicon surface enhanced Raman Spectroscopy and capacitively-coupled radiofrequency heating ,Zhenjia Wang and Lewis J. Rothberg [10] Surface Enhanced Raman Spectroscopy ,Jonathon Speed

23