RAMAN AND SURFACE ENHANCED RAMAN SPECTROSCOPY (SERS) AS A PREDICTIVE TOOL FOR MALARIA DIAGNOSIS
MSc THESIS
Kwarkye Kyei
Master Thesis May 2016
Department of Physics and Mathematics
University of Eastern Finland
Kwarkye Kyei Raman and SERS as predictive tool for malaria diagnosis , 70 pages University of Eastern Finland
Master’s Degree Programme in Photonics Supervisors Prof. Pasi Vahimaa
Ph.D. Tarmo Nuutinen MSc. Antti Matikainen
Abstract
Malaria is an epidemic that causes death mostly among children and pregnant women.
Early detection of malaria is imperative to enhance drug administration which will avert the numerous deaths that occur. A malaria detection technique should be sen- sitive, specific, cost effective and takes less time for diagnosis.
In this work, Raman and SERS was used as a technique for malaria detection by modeling it with hematin and methemoglobin in a solution of NaOH. A detection limit of 160 nM with SERS which is approximately equal to 894 parasites that occur per every microliter of infected blood sample was achieved with both hematin and methemoglobin. In a similar vain, the laser power dependence of SERS spectrum was found at low concentrations of hematin and observed that some peaks that were hidden became visible. This was indicative that the laser power should be optimally chosen in order not to burn the sample under investigation and as well not suppress the peaks. Sample preparation and spectra acquisition time was about 30 minutes.
It can therefore be speculated that SERS can be used in the early detection of malaria parasites that is the ring stage of infection. We can also predict that the SERS tech- nique can be used at malaria endemic areas due to its relatively short time of sample preparation and spectrum acquisition.
Preface
I first thank the Almighty God for his guidance and grace throughout my studies.
I would also thank Professor Pasi Vahimaa for his decision to supervise my thesis and his contributions to it. I am also grateful to my supervisors, Dr. Tarmo Nuuti- nen and Antti Matikainen (M.Sc.) for their immense contributions throughout the measurement process and thesis drafting. To all at the Department of Physics and Mathematics whose guidance and support has brought me this far i appreciate.
I thank my parents for their constant support and prayer. I also express my greatest gratitude to all and sundry whose efforts cannot be overlooked especially Awuni Emmanuel Kolog (M.Sc.) and Bawuah Prince (M.Sc.) of the Computer Science and Photonics Department respectively.
This piece is dedicated to my sons Kwarkye Asare Boamah and Kwarkye Asare Brobbey.
I will end by this popular adage in Akan that says ”when a child knows how to wash the hands, he/she dines with elders”.
Joensuu, the 20th of May 2016 Kwarkye Kyei
Contents
1 Introduction 1
1.1 Background . . . 1
1.2 Research problem . . . 4
1.3 Research objectives . . . 5
1.4 Significance of research . . . 5
1.5 Organization of the thesis . . . 6
1.6 About the researcher and his role . . . 7
2 Theory 8 2.1 What is malaria? . . . 8
2.2 Malaria parasite cycle in the human body . . . 8
2.3 Hemoglobin . . . 10
2.4 Methemoglobin . . . 11
2.5 Denaturation feature of NaOH . . . 11
2.6 Hematin . . . 12
2.7 Raman theory . . . 12
2.8 Surface enhanced Raman spectroscopy (SERS) . . . 14
2.8.1 Electromagnetic enhancement . . . 16
2.8.2 Chemical enhancement . . . 18
2.8.3 Selection rules . . . 18
2.9 Average enhancement factor (AEF) . . . 19
2.9.1 Theory of Raman spectrometer . . . 19
2.10 Malaria detection approaches . . . 21
2.10.1 Light microscopy . . . 21
2.10.2 Fluorescent microscopy and Two photon absorption . . . 23
2.10.3 Dark field microscopy . . . 23
2.10.4 Microfluidic approach . . . 24
2.10.5 Rapid diagnostic test (RDT) . . . 24
2.10.6 Polymerase chain reaction technique (PCR) . . . 25
2.10.7 Other optical imaging techniques for the study of malaria . . . 25
3 Materials and methods 27 3.1 Sample specification and preparation . . . 27
3.2 Substrate preparation . . . 28
3.3 Photo-activation of the substrate . . . 30
3.4 Analyte deposition . . . 31
3.5 Acquiring Raman spectra . . . 31
3.6 Raman spectra processing . . . 31
4 Results and Discussion 33 4.1 Raman spectra of methemoglobin and hematin . . . 33
4.2 SERS spectra of hematin and hemoglobin . . . 36
4.3 Concentration dependence of SERS enhancement . . . 40
4.4 Discussion . . . 44
5 Conclusions 48 5.1 Limitations and constraints of study . . . 49
5.2 Future Works . . . 50
Bibliography 52 Appendix A MATLAB CODES . . . 64
Chapter I
Introduction
1.1 Background
Malaria, arguably described as one of the world’s deadliest epidemics, kills an es- timated number of 584000 people and affects about 198 million people each year often children less than five years with 3.3 billion people at risk of being infected [1].
This is an indication that for every 1 minute someone dies from malaria. Malaria, is a worldwide headache since it is not only limited to people living in malaria prone regions but also travelers around the world stand a risk of this infection.
The word ’malaria’ stems from the Italian word ”mal’ aria”meaning ”bad air”
and was in 1740 introduced in English literature by Horace Walpole. The genus plasmodium is the main causative organism of malaria and has about 200 different species [2]. Human infection is primarily caused by 5 main species; they are plas- modium falciparum, malariae, ovale, vivax and knowlesi and they are transferred by the bite of female Anopheles mosquito [2, 3].
Plasmodium falciparum accounts for a majority of the worldwide deaths of ap- proximately 1 million out of 278 million reported cases in 1998 [4]. It is worth mentioning that, despite the figures given above, the epidemiology of malaria is least understood but maps for geographical distribution are readily available. It is predominant in Africa, South-East Asia and the Western Pacific regions [5, 6]. In most of the countries within these regions, one major setback is the lack of resources for malaria diagnosis.
The need to diagnose malaria and quantify the parasitic density is crucial in the control and treatment of the disease. When parasitic density (number of parasites
per microliter of blood) is not known, drug administration to a malaria-infected person and resistance to malaria drugs become a major challenge. Besides, this will eliminate the problem of inaccurate malaria cases estimated globally and give very precise reportage.
The bite of Anopheles mosquito leaves sporozoites beneath the skin due to in- gestion of blood by the mosquito. These sporozoites migrate to the liver and red blood cells (RBCs) [7]. Through molecular interactions using amino acids in protein synthesis, the sporozoites replicate into so many merozoites inside the RBC. Degra- dation of the RBC begins, leading to the release of heme a toxic substance to the parasite [8]. It detoxifies the heme by changing it into crystalline hemozoin leading to the release of free iron [9].
The conventional standard for the detection of plasmodium parasites has been observing smears in blood samples by white light microscopy [10]. The lack of trained microscopist, relatively long sample preparation time, labour intensive and expensive microscopes to observe these smears makes it difficult to detect malaria. Besides, most patients infected with plasmodium falciparum, identifying the parasites may be difficult, since the parasites may be isolated and hidden [11]. Alternatively, a more rapid way to detect malaria is the rapid diagnostic test (RDT) which relatively doesn’t require experts to operate but sometimes gives false positive and negative results [12–14]. A more sensitive way to detect and quantify plasmodium genotype is bypolymerase chain reaction technique (PCR) [15] but it takes a longer detection period and relatively costly as well.
Raman spectroscopy, a technique realized due to the interaction of light with matter is principally based on the inelastic scattering of light. Smekal in 1923 was the first person to make predictions about the inelastic scattering of light and then observed experimentally in 1928 by Raman and Krishman by focusing sunlight onto a sample using a telescope. The scattered light was collected by a lens and a system of filters whose frequency changed with incident scattered light [16, 17].
Since its inception, Raman spectroscopy has seen major applications in the med- ical field; using its principle to detect diseases, in the field of archaeology; to analyze objects of arts and antiquities, in food science; to check the adulteration of food substances, just to mention a few [18–20].
A major setback of Raman spectroscopy is the damping of the Raman signal with a scattering cross-section of approximately 10−30cm2 per molecule [17]. In the quest
to improve the Raman signal, various techniques have been explored. Prominent among these techniques issurface enhanced Raman spectroscopy (SERS), which was first described in 1974 [21]. However, its inception saw major applications in the year 1977 in Natural Science [22]. This has grown exponentially over the years in terms of research as shown explicitly in fig. 1.
1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
Years 0
200 400 600 800 1000 1200 1400 1600 1800 2000
Number of publications
A bar chart showing the number of publications on SERS between 1979 to 2015
Figure 1.1: Meta-analysis of the number of publications in web of science for the term ”surface enhanced Raman spectroscopy” as ac- cessed on February 24th, 2015.
SERS, an interdisciplinary field that touches the boundaries of all the physi- cal sciences requires developing metal nanoparticles like gold, silver or copper on a substrate which can be a metal or a silicon wafer to enhance the Raman signal by several orders of magnitude. The enhancement of the signal depends on fac- tors like, the ability of the molecules to be detected to attach themselves on the metal surface and the plasmon resonances that occurs in the metal [23]. Much has been done by playing around the enhancing factors above to obtain better Raman signals. For example, silver nanohexagonal column and gold coated naturally oc- curring butterfly wings have been developed [24]. SERS has seen applications in many fields of study including bio-sensing (disease detection), spectroelectrochem- istry (predict the behavior of molecules in different oxidation states), single molecule
sensing etc. [24–26].
1.2 Research problem
Malaria prone regions are faced with challenges of inadequate finances to acquire ex- pensive malaria detection gadgets as well as securing the services of limited technical experts to operate these devices [27]. This is a problem and has, in effect, resulted in many of the health sectors to rely on ineffective indigenous detection approaches to aid in malaria detection [27]. With this challenge, our technique in this research is relatively cheap and effective to motivate its massive usage.
Another major challenge is being able to identify the number of parasites in an infected blood samples at the early stages. This is imperative to inform the administration of drugs which will subsequently lead to its cure. However, the existing detection gadgets are not able to detect the early stages of its infection until it is aggravated [28,29]. This implies that the plasmodium in the blood samples would have to be multiplied to a certain threshold before it can be detected. Even with that, it takes a longer time for diagnosis, which is not efficient enough given the contest of highly malaria prone areas. This highlights the key relevance of our approach in this study. Our detection technique, in this work, is relatively fast with high sensitivity level.
Moreover, many of the existing techniques used for detecting plasmodium para- sites are relatively slow and requires that the blood sample undergoes preparation either by staining or drying [30, 31] which may eventually affect the outcome of the results. In its preparation process, the blood sample could be contaminated with external bodies. With this, the result may not give a true reflection of the actual state of the blood sample with respect to malaria detection. This is a problem and partly motivated the idea of this study.
In this work, we present Raman spectroscopy and surface-enhanced Raman spectroscopy (SERS) spectra of hematin and methemoglobin (non-oxygen binding hemoglobin). By reducing the concentrations of hematin and methemoglobin in a solution of NaOH, we arrive at the minimum concentration where the spectra are no more visible by employing SERS. Comparisons are made with theoretical con- centrations of plasmodium infected blood samples to predict the feasibility of SERS as a technique for early malaria detection.
1.3 Research objectives
Our first objective in this work is to simulate malaria infection using hematin and hemoglobin as model analytes and acquire the Raman and SERS spectra of each.
This will be used in subsequent research as a predictive tool in malaria infected red blood cells and healthy blood cells. If appropriate spectra is obtained we can then deduce our platform as viable for malaria detection.
After simulation and obtaining the spectra, we vary the concentrations of hematin and methemoglobin in a solution of NaOH and observe the minimum concentration where SERS spectra are not visible. This will enable us to predict the number of parasites that our system can detect at various stages of malaria infection more importantly, the ring stage of infection. Comparisons are subsequently made with theoretical parasitic concentrations in infected blood samples .
To differentiate between the spectra of hematin and methemoglobin in solution which will eventually differentiate between the various plasmodium infections. Our last objective is to arrive at a technique which will require least or no sample prepa- ration. Malaria sample preparations sometimes lead to false test results and as such a technique which requires less or no preparation is required.
1.4 Significance of research
Studies have shown that malaria prone regions are among the countries with low av- erage per capital income. Current spending on malaria control pegs around US$100 million per year [1]. This is not sufficient, yet experts have estimated in 2007 that to totally eradicate malaria, a total cost of US$2.7 billion per year which is projected to increase to US$4 billion is supposed to be borne by malaria prone countries and the world at large [27]. There has been programs to control the malaria pandemic which includes sensitizing people on the use of mosquito nets, the malaria vaccine initiative but it leaves a lot to be desired [14, 32]. It should be noted that there is a thin line between malaria control and research. With this global challenge in the full glare at our forefronts, much studies needs to be carried out on malaria diagnosis.
The wrongful administration of drugs to malaria infected patients is quite alarm- ing and this may sometimes be attributed to lack of effective detection technique [28, 29]. This may lead to the parasites still hovering in the bloodstream or at worse cause death mostly among the rural poor. Therefore a more reliable technique that
will quantify the density of the parasites in infected blood sample is important. This will enhance effective drug prescription and administration as well as eradicating drug resistant parasites.
False positive test results may be proposed due to sample preparation. When- ever blood samples undergo preparation there may be fungi infections, precipitated stain, dirt, debris from cell and bacteria which may be mistaken for malaria parasites and yields false results [30]. To avoid these infections a technique which requires little or no sample preparation is therefore needed. Likewise, as the parasitic density decreases there is a greater probability of getting false negative test by using white light microscopy [31]. Microscopic experience is therefore required to correctly inter- pret such results. It is therefore necessary to obtain a technique that can correctly detect very low parasitic densities.
Our set objectives which seek to address most of these challenges when achieved will be able to quantify parasitic density at even the ring stages of malaria infection enhancing early detection and proper drug administration. Moreover, a technique which requires less sample preparation will reduce the effect of false test result making it a more accurate detection method. The ability to make a differentia- tion between methemoglobin and hematin spectra leads to the categorization of Anophelesspeciesinfection which addresses the challenge of specificity of detection.
1.5 Organization of the thesis
The introduction chapter gives a general background of the study, its significance, the research problem and its identification. A little background about the researcher and how he is motivated on his current study is also captured.
In chapter two of this thesis, the theory as well as the works on malaria detection and how it has evolved is reviewed. This chapter also looks at the theory behind Raman spectroscopy and SERS. The causative agent of malaria, life cycle of plas- modium parasites in the blood stream is vividly discussed. Analysis are also made on the samples used in the detection process of the study.
In chapter three, the method used which involves sample preparation and fabri- cation of the SERS substrate is presented. The general experimental procedure and parameters used in the measurement process is also presented.
The results and discussions on this is presented in the fourth chapter. The
limitation of the study and what can be done in the future to overcome this limitation is also presented.
In the last chapter conclusions are made and possible measures that can enhance the experimental results are proposed.
1.6 About the researcher and his role
The researcher holds a Bachelors degree in Physics majoring in electronics from the Kwame Nkrumah University of Science and Technology in Ghana. Currently, he is studying a master degree in Photonics at the University of Eastern Finland work- ing towards completion of the program. His interest is into Biomedical Photonics, applying various techniques to study cellular and sub-cellular structures.
After completion of the Bachelor’s program he undertook a mandatory National Service in his home country Ghana at The Presbyterian Senior High School, Bom- pata. He served in the capacity of a classroom teacher where he taught subjects like Physics, Elective Mathematics and Integrated Science. After a year of service he was employed by the Ghana Education Service where he continued with his teaching.
As part of this responsibility he served as a class master where he was in charge of the well-being of assigned students related to health, social and psychological issues.
The researcher is aware of malaria infection and has been infected on several occasions and is a victim of wrongful drug administration and has been subjected to malaria-related complications. As a teacher who was once entrusted with monitor- ing the well-being of students, he knows how malaria infection and early detection hampers academic work among young and old. He is aware of the various challenges faced by health facilities in the detection and treatment of malaria especially in the remote areas like Bompata. With the current growing strength in Biomedical Photonics and its accompanied basic techniques in disease detection his interest to employ one of these techniques to detect malaria is invaluable.
Chapter II
Theory
2.1 What is malaria?
Malaria is a disease caused by infection of one of the plasmodium species. The main vector for the transmission of this virus is a bite of the female Anopheles mosquito. Of the 450 existing Anopheles mosquito species, about 70 are able to spread the plasmodium virus. Some of these species that can spread the virus includeAnopheles culicifacies, Anopheles stephensi, Anopheles fluviatilis, Anopheles minimus, Anopheles dirus, Anopheles sundaicus and many others.
TheAnopheles mosquito is an insect that responds to environmental changes like rainfall, altitude and temperature. They thrive very well in a temperature range of 20−30 oC and a humidity > 60 % [33]. Above and below these environmental conditions the vectors doesn’t survive. In spite of these environmental conditions facilitating the spread of the virus, the availability of a pool of humans already infected with the virus also aids in the transmission of the disease [33]. This is usually the case in malaria endemic areas.
2.2 Malaria parasite cycle in the human body
Whenever an infected female Anopheles mosquito sucks blood from beneath the hu- man skin, a host of sporozoites are injected into the dermis. An estimated 15−123 sporozoites [34] are deposited on the dermis which eventually migrates to the liver through the lymphatic system and red blood cells [7]. They invade the hepatocyte of the liver, a cycle that involves complex molecular interaction which uses protein to form a parasitophorous vacuole [35]. The parasites within the hepatocyte repli-
cates into new merozoite parasites which are then released from the parasitophorous vacuole to initiate infections in the blood [8].
Merozoites that are released reside in the host erythrocytes (red blood cells;
RBCs). RBC is mainly made up of hemoglobin and this hemoglobin consists of about 95 % protein [9]. During the short stay of the plasmodium virus, it degrades about 75 % [36] of hemoglobin by using it as an amino acid source aiding in protein synthesis.
Ingestion of hemoglobin is done by the parasite through the cytostone leading to the release of heme which is toxic to the virus. It detoxifies the acidic heme into a crystalline array called hemozoin by the release of excess iron. A complete cycle of the parasitic invasion is explicitly shown in Figure 2.
(a) (b)
Figure 2.1: (a) A bite of a mosquito leaving sporozoites on the der- mis of the skin (b) the parasite cycle in the human body (adapted from [37, 38]).
2.3 Hemoglobin
Hemoglobin is the protein molecule that occurs in red blood cells. The main function of hemoglobin is to carry oxygen from the lungs to other tissues of the body and in turn transport carbon dioxide from the tissues to the lungs. Basically, hemoglobin consists of four subunits of protein called globulin chains which are interconnected.
The adult human has two alpha globulin chains and two beta globulin chains. The beta globulin chain hardly occurs in infants and fetuses. This is always substituted by two gamma globulin chains. As the infant grows, the gamma globulin chains gradually metamorphose to beta globulin chains.
The globulin chains has within them an important iron containing porphyrin which is called heme. This iron is responsible for the dual transportation of air molecules in the blood stream and as well gives the blood its characteristic reddish color. Naturally occurring hemoglobin are red in color and round in shape giving the red blood cell its shape. For a disrupted shape of hemoglobin, such as those occurring in sickle celled humans, its transport through the blood system is hindered.
The quantification of hemoglobin usually requires complete counting of the blood from a blood sample. Several methods of counting exist and typical amongst them utilizes the absorption properties of the blood sample (exposing hemoglobin to cyanide to form cyanomethemoglobin) at a wavelength 540 nm to quantify it [39].
The unit of measurement of hemoglobin is grams per deciliter (gm/dL) and the hemoglobin levels in humans vary according to age and gender.
Hemoglobin can be oxygenated or deoxygenated. Oxygenated hemoglobin is formed by the binding of oxygen to the protein heme. In most cases, this bound oxygen wanders through the blood stream till it is used as an electron acceptor eventually yielding adenosine triphosphate (ATP) which aids in intracellular energy transfer. Deoxygenated hemoglobin is that which can no longer bound oxygen. Let me hasten to add that both hemoglobin types exhibit different absorption spectra which is evident at the wavelength bands of 660 nm and 940 nm. This distinct absorption spectrum is used to quantify how much oxygen is in a person’s blood.
There exist different configurations of hemoglobin normally depending on the ionic feature of the heme containing iron. A typical example is methemoglobin. Fig- ure 1.2 shows the molecular structure of hemoglobin and its corresponding biological structure.
Figure 2.2: Molecular and biological structure of hemoglobin (adapted from [40])
2.4 Methemoglobin
Methemoglobin is a deoxygenated form of hemoglobin. The normal ferrous hemoglobin has its iron in the F e2+ state. In contrast, methemoglobin has its iron in the F e3+
state and cannot bind oxygen.
In the human blood, there exist methemoglobin in small quantities of about 1 % to 2 % produced during blood metabolism. An increased concentration of methe- moglobin exceeding the percentage values stated above is an anomaly which may be due to factors like reduced cellular defense mechanisms, various pharmaceutical compounds, exposure to various environmental agents and inherited disorders. An excess of it gives the blood an uncharacteristic bluish chocolate brownish color and makes it very difficult to estimate oxygen saturation. After blood exiting the human body, it assumes this feature of blood.
2.5 Denaturation feature of NaOH
Denaturation of protein comes about when external forces are applied on proteins leading to the loss of their structure be it primary or tertiary [41, 42]. Denatured proteins have unique properties like losing its solubility to aggregation of protein
which is typically observed in boiled egg. Adult hemoglobin is always susceptible to denaturation by NaOH due to the presence of the beta globulin chains [41]. This feature is widely used in differentiating fetal hemoglobin from adult hemoglobin as the gamma globulin chains that appear in fetal hemoglobin is resistant to alkali like NaOH (Alkali denaturation test) [42]. A denatured protein has its secondary, tertiary and quartenary bonds ruptured but keeps the peptide bonds of the amino acids.
2.6 Hematin
Hematin is another form of iron containing porphyrin, that is, it has a ferric iron ion whose ligand is hydroxide. Spectroscopically and chemically, crystalline hematin and hemezoin are identical [43]. By the use of synchrotron radiation, hematin has been examined to consist of an array of dimmers which are inter-connected by a reciprocal iron-carboxylate bonds to side chains of propaionate [44]. Most anti-malaria drugs such as chloroquine binds strongly to both the monomeric and dimeric forms of heme, thereby prohibiting heme aggregation in blood. The formation of hemozoin by the malaria parasites yet remains a subject with varying explanations. One school of thought believe an aggregation of heme which is catalyzed by proteins lead to the formation of hemozoin [45]. Figure 3.1 shows the chemical structure of hematin.
2.7 Raman theory
The interaction of photons with matter leads to processes like absorption and scat- tering or sometimes no observed effect. Two main scattering events are achieved, Rayleigh scattering and Raman scattering. The elastic scattering of incident pho- tons that result in no change in frequency is the dominant form of scattering and is called Rayleigh scattering. An oscillating electric field which acts on a material causes it to vibrate at a frequency similar to the incident electric field. The particles in the material becomes polarized and behave as a small radiating dipole. The radia- tion from the dipole is measured as the elastic scattered light a natural phenomenon which determines the color of the sky. Rayleigh scattering is though observed in solids and liquids but it’s predominant in gases. It does not change the state of the material and described as a parametric process.
Figure 2.3: Molecular structure of hematin (adapted from [46]).
Raman effect is as a result of inelastic scattering of a source of light after it irradiates a sample and distorts the electron cloud leading to a temporal energy state called the virtual state. This leads to induced nuclear motion leading to the transfer of energy from the incident photon to the molecule or vice versa. For almost 106 −108 scattered photons, approximately one of these photons undergo this scattering [47]. Raman scattering is hence described as a weak process. Two forms of Raman scattering can be realized and they are Stokes and anti-Stokes scattering.
At room temperature, most molecules occupy the lowest vibrational energy level.
When a source of radiation is incident on these molecules, they are moved to virtual states and reradiated to vibrationally excited states. The energy of such scattered photons are lower than the incident photons and are called Stokes scattering. It is the dominant form of Raman scattering. On the other hand, above room tempera- ture, some molecules are already in the vibrational excited state. Upon irradiation with an incident photon they occupy the virtual states and quickly descend to the lowest vibrational energy level. This phenomenon least occur and the energy of the scattered photon is higher than the energy of the incident photon. In Figure 3, we show all the processes; Rayleigh, Stoke and anti-Stoke scattering. The states m and
n are the different vibrational states of the ground electronic level. The intensities of the Stoke and anti-Stoke scattering depends on the number of molecules at the ground electronic level and is shown explicitly by the Boltzmann equation (2.1).
Measuring Raman scattered light from a material gives important information about the substance. They give for example, information about the electronic and structural features of various complexes like the metalloporphyrin group like hemoglobin. Raman scattered radiations are due to fundamental vibrations from the molecule under investigation which can be used to construct chemical models of reaction processes. Moreover, Raman spectra is dependent on the number of molecules producing the peak, this feature can be used to quantitatively determine the concentration of a molecule.
Unlike other forms of spectroscopy like infrared spectroscopy which requires a direct relation of the incident photon with the energy difference between the ground state and excited state, in Raman spectroscopy the incident photon energy can be higher than the energy difference.
Nn
Nm
= gn
gm
exp
−(En−Em) kT
(2.1) Nn and Nm are the number of molecules in the n and m vibrational states, g is the degeneracy factor, En−Em is the energy difference between the two levels and k is the Boltzmann constant.
Raman spectroscopy being a weak scattering process has undergone revolu- tion to enhance its signal and mechanisms like tip-enhanced Raman spectroscopy (TERS) [48], resonance enhanced Raman spectroscopy (RERS) [48], coherent anti- Stokes Raman spectroscopy (CARS) [49] and surface enhanced Raman spectroscopy (SERS) [50] have been exploited.
2.8 Surface enhanced Raman spectroscopy (SERS)
SERS was first discovered by Fleischman et al. [21] in 1974 by an adsorbed pyridine from an aqueous solution onto a silver electrode corrugated by oxidation and reduc- tion processes. An enhanced Raman signal up to the order of 106 was realized as a result of the corrugation which increased the surface area of the electrode thereby increasing the number of adsorbed molecules. Later works by Jeanmarie et al. [22]
andAlbrecht et al.[51] showed that the increase in SERS intensity was not only due
Figure 2.4: Rayleigh, Stoke and Anti-Skoke’s scattering processed
to the corrugated electrodes but also on surface plasmons that exist between the metal-dielectric interface.
In most situations, metal nanoparticles are deposited on substrates like silicon or glass. Such metals include silver, gold, copper, etc. It has been shown that silver is an effective substrate [52] but others like gold also give good enhancement. The enhanced signal is not only dependent on the roughened surfaces but also a factor like the effective surface adsorption of the analyte. Metals have outer electrons bound to the positive nucleus with its electron density extending from its surface.
Stimulation of these electrons by a light beam lead to group oscillations across the surface of the metal called surface plasmons. A direct matching of momentum of the incident photons to momentum of the plasmons should be achieved. Surface plasmons normally occur in materials with a negative real dielectric constant and positive imaginary dielectric constant. The real part of the material is responsible for the scattering while the imaginary part is responsible for absorption.
To achieve enhanced Raman scattering, oscillations of these plasmons should be perpendicular to the surface plane and this can be realized by corrugating the sur- face. The valleys normally calledlightning rods caused by roughening are the peaks that causes scattering. Many theories have been proposed to support the mecha- nisms of SERS. Two main accepted theories have been used in SERS studies and
they are electromagnetic enhancement and charge transfer (chemical) enhancement.
2.8.1 Electromagnetic enhancement
The coherent oscillation of surface electrons as a result of excitation by an electro- magnetic radiation leads to theory of surface plasmon resonance (SPR). This theory is normally supported by materials with negative real and positive imaginary dielec- tric constant (metal-dielectric interface). Figure 5 shows how propagating surface plasmons and localized plasmons behave. Its propagation is perpendicular to the direction of the metal-dielectric interface and decays orthogonal to the interface due to energy conversion to heat.
Similarly when the interaction of light occurs between particles whose size is much smaller than the incident wavelength, localized surface plasmons (LSPR) are generated. The plasmonic oscillation is around the nanoparticle and the frequency of oscillation is called LPSR. This can be explained as the basis for electromagnetic enhancement in SERS.
Figure 2.5: Schematic diagrams of (a) localized surface plasmon (b) propagating plasmon (c) frequency dependence on surface plasmon polariton (adapted from [37]).
The electromagnetic enhancement involves a direct interaction between the ana-
lyte and the surface plasmons by adsorbing it onto the metal surface. Its description is based on the metallic sphere model which is not an entirely correct model but can be used to describe simple processes [53]. This is because there are several ag- gregations of these spheres that gives an enhanced SERS. The field around a sphere which is irradiated with an electric field can be solved using Maxwell’s equations by employing the quasi-static approximation. The resulting solution is given by
Er =E0cosθ+gb3 r3
E0cosθ (2.2)
where Er is the electric field measured from a distance r from the sphere surface,b being the radius of the sphere,θis the angle as measured with respect to the electric field and g is a constant which is related to the dielectric constants. The equation relating g to the dielectric constants is given by
g = ǫ(ωL)−ǫm
ǫ(ωL) + 2ǫm
!
(2.3) ǫ and ǫm are the dielectric constants around the sphere and metal sphere of the medium, whereas ω is the frequency of the incident radiation. It can be inferred from the equation 2.3 that when the denominator becomes very small a large g value is realized. Due to the dependence of the dielectric constant of the metal nanoparticle on wavelength, g determines the resonance condition of the particle.
For ǫm value of −2 and ǫ value of 1 this condition is met and at this value of g, the surface plasmon around the molecule absorbed by the metal surface increases the total field around it. Withg’s dependence on dielectric constants and excitation frequency explicitly shown by equation (2.3) and direct substitution of g value into (2.2) it can be deduced that the total electric field is maximum at the perpendicular surface than to the parallel surface.
For molecules adsorbed on a metal surface, the enhancement is not uniformly distributed around any stand-alone particle but enhancement can exist between touching particles or a cluster of particles. The parts of the particle which remains very active after excitation with a laser frequency are described as hot spots whereas some parts remain inactive. The plasmon frequency is dependent on particle size. In general contributions like frequency, size of particle and shape, particle organization as clusters contribute greatly to the whole SERS mechanism.
2.8.2 Chemical enhancement
Chemical enhancement, also known as charge transfer, requires the transfer of charges between the analyte and the metal surface leading to the formation of a bond. There is hence a surface specie created between the analyte and the metal surface leading to polarizability of the molecule as a result of interactions with the electrons. These new bond formation leads to the creation of electronic states which are believed as resonant intermediates in the Raman scattering. The opposite is rather seen in the electromagnetic enhancement mechanism. As in the chemical enhancement, the radiations happen within the metal surface. The hole that is transferred to the metal from the analyte to form the bond is reradiated after exci- tation with an incident light source leading to the Raman process.
There is a thin distinction between the chemical enhancement mechanism and electromagnetic enhancement mechanism though there are evidence for the exis- tence of both theories. Chemical enhancement normally involves molecules that are directly attached to the metal surfaces and hence can be described to occur around a monolayer interface. In the vast majority of the mechanisms that occur, electro- magnetic enhancement contribution is greater than the contribution from chemical enhancement.
2.8.3 Selection rules
The interpretation of SERS spectra is relatively a difficult one since some peaks that are available in normal Raman spectra may not be visible. The concentration of the analyte dependence on the intensity may be nonlinear as depicted in some molecules like pyridine. This is also the usual case in molecules with a center of symmetry.
Due to this anomaly, there are proposed selection rules for the interpretation of such spectra. These selection rules best describes electromagnetic enhancement mechanisms than chemical enhancement. This is because there is no clear definition of the nature of the species formed between the adsorbate and the surface in chemical enhancement.
The existence of these selection rules have proven to be effective in determining the orientation of some molecules. It poses the advantage of differentiating Raman’s and SERS spectra. Conversely, these selection rules pose problems in the interpreta- tion of SERS spectra due to the appearance and disappearance of some peaks. SERS
can enhance Raman scattering by a factor of 106 which is normally attributed to the surface plasmon and other factors. Another factor which sometimes contribute to this enhancement may be contaminants which strongly binds to the metallic surface.
2.9 Average enhancement factor (AEF)
The degree of enhancement offered by SERS since its inception has been of impor- tance to the detection of diseases and its related areas [20,24,54]. This estimation of SERS enhancement is arguably a more practical way of characterizing the platform especially in practical applications like disease detection. The enhancement factor due to dependence on factors like substrate, analyte and the excitation wavelength has hardly been able to be used to quantify signal strength [55]. Due to the lack of concrete definitions to enhancement, there has been a proposal by Le Ru E. C. et al. to discuss the possible definitions for SERS enhancement factors [56].
These proposals are based on situation of the analysis and are enumerated as (i) the single molecule enhancement factor (SMEF) (ii) the SERS substrate point of view and (iii) analytical chemistry point of view. Herein, we focus on the analytical Chemistry point of view description in our analysis. This is because it gives a straight forward analysis of enhancement and tends to reduce emphasis on intrinsic characteristics of the substrate [56]. It is primarily concerned with how much signal can be generated from a SERS platform. Generally it is expressed as
AEF = Iλ,Augumented
Iλ,Ref
× PAugumented
PRef
× CRef
CAugumented
(2.4) where (Iλ,Augumented/Iλ,Ref) is the Raman intensity ratio at a wavenumber or wave- length shift, (PAugumented/PRef) is the excitation powers and (CAugumented/CRef) is the concentration in enhanced and reference measurement. The equation 2.4 is ideal for similar experimental conditions where either the concentration of the analyte or the excitation power is varied. One disadvantage of this equation is that, it doesn’t really characterize the number of molecules that are adsorbed on the substrate.
2.9.1 Theory of Raman spectrometer
The Raman spectrometer is basically made up of four main components, namely;
an excitation source, sample illumination system and light collection optics (mi- croscope), wavelength selector (filter) and detector. These components are shown
in Figure 3.4. Its principle of operation is based on the Raman effect. When a monochromatic light is incident on a sample, there is scattering of light which emerges out of this interaction. These scattered light may have an energy less or greater than the incident radiation. The light source that is used in Raman spec- trometers has evolved from mercury lamps in its inception to laser sources which provide a stable and intense beam of radiation [57]. Light from the light source are confined by a focusing optics to excite the sample.
Figure 2.6: A simplified optical layout of a typical Raman spectrom- eter(adapted from [58]).
Filters are used to separate the relatively weak Raman lines from Rayleigh scat- tering and sometimes to correct laser aberrations. For a single laser beam, band pass filters are used whereas a combined notch filter and grating monochromator are used in dispersive instruments. Other filters like triple grating monochromator, super notch filters, edge filters, rejection filters and holographic filters have all found applications in Raman spectroscopes [57, 59]. After filtering, a spectrograph is used to separate the scattered light into its wavelength components with the help of a diffraction grating element.
The signal is then incident on a detector which measures the intensity at each
wavelength representing the Raman spectrum. Early spectrometers were equipped with thermoelectrically cooled photomultiplier tubes (PMT) and photodiode array detectors [57]. Advances in instrumentation has led to the replacement of these devices with relatively sensitive detectors such as charge-transfer detectors (CTDs), charge-couple detectors (CCDs) and charge-injection detectors (CIDs).
2.10 Malaria detection approaches
This section presents the various methods that have been employed in detecting plas- modium parasites in blood samples. We look at the evolution of these techniques with their accompanying setbacks, thereby predicting a more reliable technique for diagnosing malaria. A detection technique, should be sensitive (able to identify maximum number of parasites/microliter of blood), specific (identify the various plasmodium infections), accurate (gives no or few false negative and positive tests), fast (results can be obtained in less than 30 minutes) [60, 61]. Moreover, the results should be free from confounding factors (patient characteristics does not affect accu- racy), a more portable instrument should be used (operates from battery and should be less than 2 kg), requires minimal experience to operate, very cost efficient (tests are run at less than 30 cent per test) and able to adapt to new genomic markers as and when they become available [60–62].
2.10.1 Light microscopy
The conventional method of malaria parasite detection from blood sample is often carried out by staining the blood and observing smears under the microscope [63].
The blood samples are normally obtained from a patient’s finger and then collected using an anticoagulant like EDTA-coated tubes meant to be used in a short time [63].
This is because, a long stay of the blood sample in the anticoagulant tube leads to the metamorphoses of the blood through so many sexual stages [64].
For diagnosis, 8 blood films should be prepared, 4 being thin blood films and 4 being thick blood films. Where sensitivity and quantification of parasitic density of the test is desired, thick films are used, whereas thin films are ideal for specificity of the test. There are many techniques in staining the blood sample which includes, Giesma stain, Zenner stain, Field stain, HRP-II antigen detection etc. Under light microscopy, Giesma and Field staining are mostly recommended [64]. For a relatively
short time diagnosis, the rapid Field staining is used but should normally be backed by Giesma stained thin films. The rest of the prepared films are used when there are problems with the diagnosis.
Examining thick films has proven to be an efficient way of diagnosis compared to thin films due to a large examination volume with a sensitivity of about 44 fold [63] better than thin blood film. However, Dowling and Shute [65] found that about 60 % of the parasites were obscured in thick film examination and was even higher at very low parasitic densities. The situation was even worse for plasmodium falciparum gametocytes rising exponentially to 86 % missing parasites [65].
There exist several methods in quantifying malaria samples in thick blood films.
One of such methods is by multiplying the white cell count by a standard value of 8000/µl. Initially, parasites are counted and for every 200 leucocytes counted, it is multiplied by a value 40 and this gives the number of parasites per µl of blood sample [63]. This method is widely employed if the microscopist is interested in the intensity of the infection. Alternatively, a very thick smear can be prepared with a known volume (0.28 µl) of blood. The smear is dried, Giesma stained and the parasites on the smear are counted. After which the total parasites counted is multiplied by 3.33 which gives the number of parasites/µl [63]. Similar counting can be done using high power microscopes of magnification around ×100 and also in thin blood smears.
Conventional light microscopy has been the major technique for malaria diag- nosis but it faces major challenges especially in malaria endemic areas. Typical of such challenges is that it is time consuming. It takes approximately an hour of sample preparation time [66]. In interpreting results, experienced microscopist with considerably higher expertise are needed.
Parasitic size of plasmodium parasites ranges from 1 µm to 15 µm requiring magnification of 500 to 1000 × to be observable by the human eye. Though most modern microscopes satisfy this range of magnification, there is a predominant color aberration which increases linearly with magnification affecting the interpretation of the results. Infections from parasites from plasmodium falciparum normally results in the parasites hiding and are not present in peripheral blood. One may interpret results from an infected blood as negative. Additionally, the lack of staining ma- terials and microscopes coupled with reading large data during malaria outbreak, which increase the room for error, is a major challenge.
2.10.2 Fluorescent microscopy and Two photon absorption
Fluorescent microscopy requires the use of a fluorochrome to stain nucleic acids of the parasites within the sample and then examined under a fluorescent microscope.
Such fluorochromes may include Acridine-orange (AO) and benzothiocarboxypurine (BCP). Baird et al. [67] and Kawamoto et al. [68] proposed the quantitative buffy- coat (QBC) and Kawamoto Acridine-orange techniques respectively which uses Acri- dine for staining. A third technique which uses benzothiocarboxypurine as stain was proposed by Makler et al. [69]. All these are readily available as commercial kits on the market. They offer the advantage of being sensitive and specific whenever thick blood smears are under study.
Two photon absorption and third harmonic generation has also been employed to image hemozoin in live RBCs [70] and has proven to be very efficient with its high signal to noise ratio and specificity. These phenomena are purely non-linear processes which require a high intense pulsed laser so that the sample under examination will have a higher probability of absorbing two or more photons at a time. Pulsed lasers being expensive and makes this technique not ideal for malaria diagnosis
Technically, these methods are demanding, requiring special expertise and equip- ments (fluorescent microscopes with high intensity halogen lamp) and as well is un- able to differentiate between the plasmodium spp. AO is hazardous and has to be disposed off with care.
2.10.3 Dark field microscopy
Dark field microscopy uses the scattering properties of the various pigments of an infected blood to detect the presence of the parasites after illumination. This tech- nique is ideal for highly pigmented malaria parasites and has been used since 1930 to detect infected RBCs. Jamjoom [71] in 1982 detected plasmodium parasites in an unstained blood sample using dark field microscopy though rapid and sensitive requires higher expertise. Moreover, scattered light may not only be due to the pig- ments but rather other small particles in the sample. In recent development, dark field microscopy coupled with Resonance Raman microscopy has been used to image infected RBCs [72]. With a major setback of scattering contrast, a more enhanced method that integrates polarization microscopy has been proposed by Wilson et al. [73].
2.10.4 Microfluidic approach
Another emerging field for the diagnosis of malaria has been how to use microfluids normally referred to as lab-on-a-chip to detect the parasites. Very cost efficient, requiring little or no experience on operation and its ability to detect several diseases makes it appropriate as a tool for detection of malaria parasites in remote areas.
This technology has not seen major utilization but offers a high potential in malaria detection as compared to the other techniques.
The Science of microfluidics requires the manipulation of complex fluids in chan- nels of very small sizes in the order of micrometers. It therefore controls the concen- trations of substances within the microstructure due to laminar flow with respect to space and time. Whenever malaria parasites invade the erythrocyte, its membranes electrical conductivity changes and this property has been explored in sorting and discriminating infected RBCs and normal RBCs [74]. Microfluidic is broadly catego- rized into four main subdivisions, namely, molecular analysis, biodefence, molecular biology and microelectronics. Microfluidic channels are able to mimic capillary blood vessels within the human body and device fabrication can be done for many useful applications.
In recent works,Rathod et. al.[75] has developed a laboratory-based microfluidic device that is very useful in laboratory setting because of the inherent microscopy properties to study the pathogenesis of malaria. They predict that such devices can also be developed to be utilized in the real field and can be custom-designed by integrating possible detector systems which will eventually eliminate microscopy as a complement to quantifying parasitemia [76]. The works byChoi et. al. has shown that not only microfluidic devices are being developed but also biosensors that are able to detect nucleic acids and proteins has been used [77].
2.10.5 Rapid diagnostic test (RDT)
RDTs work on the principle of immunochromatography which involves the move- ment of a liquid across the surface of a nitrocellulose membrane. Almost all RDTs operation is based on this principle and has seen tremendous improvement since its inception in the late 20th century. The target is always the parasitic antigen which captures the monoclonal antibodies through laminar flow. Whenever plas- modium falciparum infected blood samples are being detected, a specific antigen
called Histidine-Rich Protein 2 (HRP-2) is used as the target. In other forms of plasmodium infection, parasite Lactate dehydrogenase (pLDH) is the antigen used.
Some commercial kits carry both antigens making a distinction between p. fal- ciparum infections and other plasmodium species. RDTs diagnosis are quick to perform requiring little or no experience and sensitive as well [78].
The major challenge in the diagnosis of malaria has been how to minimize cost.
A cost effective technique should be relatively lower than the cost in gold standard microscopy detection. Microscopy cost U.S.$0.12−U.S.$0.40 per test whereas RDT test in developing countries isU.S.$0.55−U.S.$1.50 making it comparatively expen- sive [79]. Moreover issues of specificity of detection and environmental conditions like temperature and humidity affects the performance of the kit [79]. Manufacturers of RDTs normally recommend a temperature of 4o−30oC and a humidity of <70
% as the optimum conditions for which the kit works efficiently. However, these conditions are difficult to be met in tropical Africa.
2.10.6 Polymerase chain reaction technique (PCR)
The PCR technique is one of the molecular methods used in detecting nucleic acids peculiar to plasmodium parasites and has been in existence since 1984 [80]. PCR is very sensitive being able to detect parasites at very low parasitaemia of < 5 parasites/µl by amplification of malaria genome and specific as well [12]. Most malaria infected persons may have some parasites sequestered in the capillaries even after treatment which can later be released into the bloodstream. Such parasites are detected by the PCR technique. Nowadays, efforts are being made to incorporate microfluidic approach and PCR which will tend to compensate for the shortcomings of both [81]. This technique has been used in malaria endemic areas like Vietnam and has proved to be a viable tool for detection [82].
This is relatively an expensive technique requiring special equipment. Besides, it require experts to operate the equipment since multiple steps are involved in its operation. In some instances, false positive and negative tests are recorded during diagnosis due to presence of PCR inhibitors already in the blood.
2.10.7 Other optical imaging techniques for the study of malaria
Since most conventional microscopes are limited in terms of resolution due to diffrac- tion, much effort has been made to bring the spatial resolution to a nanometer scale.
Such imaging techniques include but not limited to stimulated emission depletion (STED), stochastic optical reconstruction microscopy (STORM), photo-activated lo- calization microscopy (PALM) and structured illumination microscopy (SIM). SIM has been used in detecting malaria in infected RBCs with an extremely high reso- lution [83].
Due to the dual refractive indices exhibited by merozoites and hemozoin during their mature stages and hence described as birefringent, polarization microscopy has been used to detect them in a flow cytometer [84, 85].
Several microscopy techniques like Forster resonance energy transfer (FRET), fluorescence lifetime imaging microscopy (FLIM), differential interference contrast microscopy (DIC), quantitative phase matching, refractive index tomography just to mention a few have all been applied in the field of malaria research [86–90].
Chapter III
Materials and methods
This chapter presents the materials and methods used in our detection technique.
Preparation of the methemoglobin and hematin analytes in both Raman and SERS measurement is explicitly highlighted. The substrate preparation as well as how the spectra was acquired and processed is also looked at.
3.1 Sample specification and preparation
Already prepared hemoglobin (model number: H2500) through centrifugation of blood corpuscles and lighter plasma components was obtained from SIGMA − ALDRICH. Since hemoglobin is oxidized in air this sample is methemoglobin.
This product was specifically obtained from bovine blood (lyophilized powder). It is reddish brown in color and assumes a powdery nature. About 20 mg or a little above this was soluble in 1 mL of water. It has a dark red-brown color in solution.
It has a net 0.25−0.35% of Iron (Fe) and 68% of water. Its storage temperature is 2−8oC [91].
Similarly, porcine hematin that is hematin from pig (Product number: H3281), which was dark blue to black and in the powdered form was used in our study.
About 25 mg/mol dissolves in 1 M NaOH and assumes light green - brown to black color in solution. The infrared spectrum of hematin conforms to its structure. It consist of 63% of water and >8.4% of anhydrous Iron [91].
Before acquisition of the Raman spectra, single crystals of the samples (methe- moglobin and hematin) were taken and placed on a very clean glass slide. They were manually granulated into fine mirror structures where the sizes of the particles approached the micrometer scale. Less concentration of the samples were used in
the Raman spectra acquisition.
Analyte solutions for SERS spectra acquisition were prepared by dissolving solid crystals of methemoglobin and hematin in NaOH solution, bearing in mind the molecular weight of hemoglobin was about 101 times that of hematin. This was considered in the amount chosen of these samples that were dissolved in 0.01 M of NaOH. Our choice of NaOH as a solvent was due to the insolubility of hematin in water. Besides, the denaturation properties as described in Section 2.5 offered by NaOH was another factor that led to its choice. An initial 16µM solutions of both samples were prepared. The concentration was subsequently diluted to solutions of concentration 1.6 µM and 160 nM. Figure 4.1 shows the substrate preparation process.
3.2 Substrate preparation
The SERS substrate fabrication described here and used in our study is based on the method in [52].
A 60 mL each of silver nitrate (AgNO3) and sodium chloride (NaCl) with a con- centration of 50 mM was poured into graduated beakers. They were then diluted with 540 mL of deionized water to 600 mL yielding a concentration of 5 mM. A silicon wafer 2 inches in diameter and 0.25 mm thickness with a crystalline orienta- tion of (1,0,0) was continuously immersed into these already prepared solutions in a cyclic manner under normal laboratory conditions. A home built motorized device was used in the preparation process. The substrate was attached to suction cap to hold it firmly in the fabrication process. A magnetic stirrer (Heidolph MR 3001K) was used to stir the solutions throughout the preparation process. After a complete chemical reaction between the Ag+ and Cl− ions under normal laboratory condi- tions, AgCl crystals were formed on the surface of the wafer. A total of 50 cycles which lasted approximately 15 minutes was used. The prepared substrate was then rinsed in double distilled water for 5 minutes and dried using a homemade spinner.
It should be noted that the growth of the AgCl crystals is dependent on factors like concentration of the solutions, the number of cycles and cycle duration. After the growth of the AgCl nanoparticles, the wafer was then ready for photo-activation.
Scanning electron microscope (SEM) images showed that after 50 cycles, the crystals were distributed evenly and formed a wider bell shape. At this number
Figure 3.1: A simple setup of the substrate preparation process (adapted and redrew from [58]).
(a) (b)
Figure 3.2: SEM images of the crystals (a) before (b) after pho- toreduction with magnification of 25000
of cycles there seemed to be contributions of enhancement from both smaller and larger AgCl crystals from around 30 to 150 nm in diameter. Figures 3.2 shows SEM images of the crystals before and after photoreduction.
3.3 Photo-activation of the substrate
The AgCl coated wafer was not SERS active and has to be reduced to the metal Ag. Since SERS mechanism is principally based on plasmonic effect as a result of contributions from metal nanoparticles, AgCl has to be reduced to silver to bring about this effect. The photo-reduction was achieved by making a small spot of di- ameter 100µmon the prepared AgCl crystal. A laser light source with an excitation wavelength of 514 nm, excitation power of 5 mW, was shun on the wafer for about 2 minutes using the 5X objective lens with a numerical aperture of 0.40. The contin- uous exposure of the excitation source sees to the release of chlorine leaving behind the silver nanoparticles.
Figure 3.3: A laser being used to reduce AgCl to Ag.(adapted from and redrew from [58])
3.4 Analyte deposition
Prior to the acquisition of the Raman spectra, three main sample manipulation techniques were exploited. The first technique is incubation, which requires the deposition of the analyte droplet on the surface of the wafer and left to partially dry for some time. The liquid part which does not dry is blown away using air. This technique suffers a setback of the possible blowing away of the analyte molecules which may not have attached themselves properly with the silver nanoparticles. The second method require measuring through the deposited analyte without any drying.
It was observed that the analyte formed a lens-like structure on the surface of the wafer. This made focusing of the excitation source on the analyte very difficult as well as scattering by this lens-like feature.
A third technique which was employed in the acquisition of the SERS spectra in this work required the deposition of the analyte on the silicon wafer and left for about 5 minutes to dry completely. About 5 µL of the analyte was used to obtain the spectra in each measurement. There was an observed attachment of hematin to the silver nanoparticles than to methemoglobin. It should be noted also that NaOH did not react with either the silicon wafer and or the silver nanoparticles.
3.5 Acquiring Raman spectra
A Renishaw inVia Raman microscope which is controlled by a wire 3.4 software was used in the acquisition of the spectra within the wavenumber range of 100−1700 cm−1. In normal Raman measurements, Raman spectra was acquired using an excitation wavelength of 514 nm, a laser power of 0.05 mW, 100X objective and a time of 60 s. In acquiring the SERS spectra, an excitation wavelength of 514 nm, a laser power of 0.05 mW using the 20X objective in a time 10s was employed. The laser power was increased from 0.05 mW to 0.5 mW in other to observe possible enhancement of Raman signal of hematin.
3.6 Raman spectra processing
The spectra, after acquisition were processed and analyzed using MATLAB (Re- lease 2015b; The Mathworks Incorporation, USA). The file which was obtained as an SPC file format (.spc) file from the Wire 3.4 software consisted of spectra from
Figure 3.4: A simple setup of the Raman acquisition process (adapted from [58]).
five different points from the wafer. The built in average function was initially used to average all the five spectra. The spectra was subsequently denoized using the
’wden’ function, a one dimensional denoizing tool which is based on the wavelet decomposition and reconstruction method [92]. A threshold selection rule ’rigrsure’
which is based on the Stein’s principle of unbiased risk was selected. A level de- pendent estimation of noise was used to rescale the decomposed spectra using an orthogonal wavelet level 10. The soft threshold which is a preprocessing tool was used to reduce the background in the spectrum
The denoizing procedure can broadly be summarized as decomposition of the spectra, detailed coefficient thresholding and reconstruction of the spectra. The spectra was further smoothened using the ’sgolayfilt’ function which applied a Savitzky- Golay FIR smoothing filter to the data. Savitzky-Golay filters are preferred to stan- dard averaging FIR filters because they filter the signal high frequency components alongside the noise keeping the most important high frequencies. A polynomial of the order 4 was chosen with a frame size of 9. Since less fluorescence was observed in the original spectra, no fluorescence baseline was set.
Chapter IV
Results and Discussion
This chapter reports on Raman and SERS spectra of methemoglobin and hematin.
By varying the concentration of these analytes in NaOH solvent, a SERS platform is used to predict how sensitive our detection technique can be. Other factors like power of laser dependence on intensity of the Raman peaks is also looked at as well as the feasibility of our SERS system for malaria detection.
4.1 Raman spectra of methemoglobin and hematin
Methehemoglobin crystals were flattened to fine mirror crystals. The Raman spectra of the relatively higher concentrated fine crystals were acquired under an excitation wavelength of 514 nm. This excitation wavelength was chosen because the analytes were non-fluorescent. Figures 4.1 show the Raman spectra for a wavenumber range of 100−1700 cm−1.The assignments proposed to the peaks in the resulting spectra are based on the labelling scheme developed by Abe et al. [93].
It is evident from Figure 4.1 that there are four distinct peaks at the low wavenumber region of 100 −1190 cm−1. The peaks at 677 cm−1 and 757 cm−1 is attributed to porphyrin bands ν7 and ν15 respectively [94, 95]. They are modes of the pyrrole breathing and symmetric pyrrole deformation. Within the same low wavenumber region, there exist bands at 1168 cm−1 and 1125 cm−1 which are as- signed to the asymmetric pyrrole half-ring stretching vibrations ν30 and ν22 [50, 96].
In proteins the stretch is due to C-N and C-C bonds.
The work of Salmaso et al. [97] proved that there exist three main bands in the wavelength region 1200-1300 cm−1. The appearance of some of these bands within this region is dependent on the excitation wavelength of the laser source used. This
(a) (b)
Figure 4.1: (a)Shows the characteristic fingerprints of Raman within the wavelength range of 100 −800 cm−1 whereas (b) gives the spectrum between 800−1700 cm−1 of methemoglobin with a laser power of 0.05 mW and spectra acquisition time of 60 s.The positions of the spectral fingerprints are indicated by arrows.
region is called the methine C-H deformation region and there appears a single distinct peak for methemoglobin at 1240 cm−1 with an excitation wavelength of 514 nm. This peak is assigned toν13 or ν42 and is due to threonine rocking of CH3.
The spin state marker band region, which falls within the range of 1650−1500 cm−1, has three main bands appearing in this region for our excitation wavelength.
These bands are 1640, 1586 and 1564 cm−1 and are assigned to ν10, ν37 and ν2
respectively. The band assignment atν10is due to proteins such as amide I whereas the assignments ν37 and ν2 are due to heme. There exist a very feint amino acid contribution due to the phenylalanine mode at 1005 cm−1. These Raman features are in accordance with what has been reported in literature, though there are slight shifts in the peaks which is dependent on factors like temperature, preparation of the crystals as well as the excitation wavelength of the laser.
Interestingly, the bands which appeared at 677 cm−1 and 558 cm−1 are normally
observed in oxygenated hemoglobin. This is as a result of iron-oxygen vibration resulting from Fe-O-O stretching. It can be inferred that these peaks were observed in methemoglobin spectra because of the oxidation of iron with air molecules during sample preparation. These bands are though very less appearing.
Figure 4.2 shows the Raman spectra of hematin which was obtained by finely granulating its crystals into fine mirrors as was done in the methemoglobin crystals.
The same measurement conditions employed in methemoglobin was used in the acquisition of the spectra of hematin.
(a) (b)
Figure 4.2: (a)Shows the characteristic fingerprints of Raman within the wavelength range of 100−800 cm−1 whereas (b) gives the spectrum between 800−1700 cm−1 of hematin with a laser power of 0.05 mW and spectra acquisition time of 60 s.The positions of the spectral fingerprints are indicated by arrows.
Relatively strong Raman peaks were observed at 757 cm−1, 1372 cm−1, 1569 cm−1 and 1628 cm−1. Similar observations also occurred for methemoglobin at 757 cm−1, 1375 cm−1, 1586 cm−1 and 1640 cm−1. The peaks 1375 cm−1 and 1372 cm−1 are similar peaks and the slight shift may have arisen because of factors like temperature and sample preparation. The comparatively intense peak at 1586 cm−1,
