USF Libraries
USF Digital Collections

Hemocompatibility assessment of 3c-sic for cardiovascular applications


Material Information

Hemocompatibility assessment of 3c-sic for cardiovascular applications
Physical Description:
Schettini, Norelli
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:


Subjects / Keywords:
Cubic silicon carbide
Platelet adhesion
Cell proliferation
Dissertations, Academic -- Electrical Engineering -- Doctoral -- USF   ( lcsh )
non-fiction   ( marcgt )


ABSTRACT: The hemocompatibility of crystalline Silicon Carbide (SiC), in its cubic form (i.e., 3C-SiC), has been evaluated and compared to Silicon (Si), the leading material in biosensing applications. Silicon carbide (SiC) is a hard, chemically robust material, very well suited for harsh environment applications, and has been suggested to have very good biocompatibility. Additionally, SiC in its amorphous form, has been used as a coating for medical implantable devices such as bone prosthetics and cardiovascular stents. However, assessment of single crystal 3C-SiC for cardiovascular applications has not been reported. In this research we have studied the interactions of single crystal 3C-SiC with platelets and human microvascular endothelial cell (HMVEC) to assess the degree of hemocompatibility of 3C-SiC. The more hemocompatible a material is, the less platelet adhesion would be expected. Using fluorescence microscopy higher platelet adhesion was statistically observed on Si than on SiC. In addition 3C-SiC surfaces showed less platelet reactivity, measured by the degree of platelet adhesion, aggregation and activation, with mostly circular morphology of adhered platelets while Si showed an elevated presence of non-activated (Circular) platelet clumps. Additionally, HMVEC proliferation assessment suggest that 3C-SiC performs comparably to high attachment culture wells with enhanced proliferation, without affecting cell morphology. These results suggest that 3C-SiC is a promising candidate for applications in the blood stream due to its low thrombogenic characteristics and good hemocompatibility
Dissertation (Ph.D.)--University of South Florida, 2009.
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Norelli Schettini.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0003266
usfldc handle - e14.3266
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003266
XX9999 (Online)
1 100
Schettini, Norelli.
0 245
Hemocompatibility assessment of 3c-sic for cardiovascular applications
h [electronic resource] /
by Norelli Schettini.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Includes vita.
Dissertation (Ph.D.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: The hemocompatibility of crystalline Silicon Carbide (SiC), in its cubic form (i.e., 3C-SiC), has been evaluated and compared to Silicon (Si), the leading material in biosensing applications. Silicon carbide (SiC) is a hard, chemically robust material, very well suited for harsh environment applications, and has been suggested to have very good biocompatibility. Additionally, SiC in its amorphous form, has been used as a coating for medical implantable devices such as bone prosthetics and cardiovascular stents. However, assessment of single crystal 3C-SiC for cardiovascular applications has not been reported. In this research we have studied the interactions of single crystal 3C-SiC with platelets and human microvascular endothelial cell (HMVEC) to assess the degree of hemocompatibility of 3C-SiC. The more hemocompatible a material is, the less platelet adhesion would be expected. Using fluorescence microscopy higher platelet adhesion was statistically observed on Si than on SiC. In addition 3C-SiC surfaces showed less platelet reactivity, measured by the degree of platelet adhesion, aggregation and activation, with mostly circular morphology of adhered platelets while Si showed an elevated presence of non-activated (Circular) platelet clumps. Additionally, HMVEC proliferation assessment suggest that 3C-SiC performs comparably to high attachment culture wells with enhanced proliferation, without affecting cell morphology. These results suggest that 3C-SiC is a promising candidate for applications in the blood stream due to its low thrombogenic characteristics and good hemocompatibility
Advisor: Stephen E. Saddow, Ph.D.
Cubic silicon carbide
Platelet adhesion
Cell proliferation
Dissertations, Academic
x Electrical Engineering
t USF Electronic Theses and Dissertations.
4 856


Hemocompatibility Assessment of 3C SiC for Cardiovascular Applications by Norelli Schettini A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Phil osophy Department of Electrical Engineering College of Engineering University of South Florida Co Major Professor: Stephen E. Saddow, Ph.D. Co Major Professor: Mark Jaroszeski, Ph.D. Anne Curtis, M. D. Karl E. Muffly, Ph.D. Andrew Hoff, Ph.D. Jing Wang, Ph.D. Date of Approval: October 30, 2009 K eywords: cubic silicon carbide, thrombogenicity platel et adhesion, circularity, cell p roliferation Copyright 2009, Norelli Schettini


Dedication To those that in the lack of motivation provided me with inspiration.... To the ones that kept me awake, sane. To the ones that didn't give up on me. To the ones that nourished my dreams. To the ones that helped me find the path. To the ones that were with me when I lost the path... ... when I lost the patience. To those that in the lack of inspiration provided me with strong motivations...


Acknowledgments In the first place, I would like to express my gratitude to Dr. S. Saddow, who has supported this project intellectually, morally and financially, he offered me the opportunity to work in his group, started me in the area of biotechnology and provided me with a sense of belonging, motivations, a working environment and infinite hours of academic discussion that resulted in this document. My gratitude also goes to Dr. M. Jaroszeski, he accepted to be my co major professor, and provided the right tools and necessary support to develop this research. I am also indebted to my committee members Dr. A. Curtis, Dr. K. E. Muffly, Dr. A. Hoff and Dr. J. Wang, it has been an honor to work with them and they have provided very valuable suggestions th r ough the course of this research. Next, I would like to thank especially Mr. L. West and Dr. K. E. Muffly who became pivotal for the development of this research; I owe most of my biotechnology knowledge to them Additionally, I would like to thank the FCo E BITT fellowship and Dr. T. Vickerson for all her support.I would like to acknowledge Dr. L. Ondrovic and the staff of the Department of Comparative Medicine (USF Vivarium ) for our blood and tissues supply Finally, I would like thank my friends and colle agues for their valuable help and support: Dr. Coletti, Dr. Reyes, Dr. Walker, Dr. Frewin, Dr. Benjamin, Dr. Short, C. Locke, I. Haselbarth, A. Oliveros, E. Murphy, our visitors Dr. A. Severino and R. Anzalone (AFM) to R. Connelly, J. Rey, J. Jimnez a nd e specially to J. Elliott to whom I owe the survival of my cell cultures.


i Table of Contents List of Tables ................................ ................................ ................................ ..................... iii List of Figures ................................ ................................ ................................ .................... iv Abstract ................................ ................................ ................................ ............................. vii Chapter. 1: Introduction ................................ ................................ ................................ ...... 1 1.1 Research Objective and Motivation ................................ ............................. 2 1.1.1 Thrombus Formation ................................ ................................ ....... 4 1.2 SiC Properties ................................ ................................ ............................ 10 1.2.1 SiC a Promising Material for Biosensing Applications in the Blood Stream ................................ ................................ ............ 12 1.3 Summary and Dissertation Organization ................................ ................... 21 Chapter. 2: Biocompatibility of Materials ................................ ................................ ........ 22 2.1 Biocompatibility of Materials with Blood: Hemocompatibility ................ 23 2.2 Standard ISO 10993 4 ................................ ................................ ................ 25 2.3 Perspectives in Hemocompatibility Assessment ................................ ....... 27 2.3.1 The Platelet Adhesion and Activation Assessment Perspective ................................ ................................ ..................... 30 2.3.2 The Protein Adsorption Assessment Perspective .......................... 32 2.3.3 The Endothelial Cell Proliferation Perspective .............................. 33 Chapter. 3: Platelet Adhesion and Activation ................................ ................................ ... 35 3.1 The Blood and its Elements ................................ ................................ ....... 35 3.2 Platelet Adhesion and Activation Assessment Protocol ............................ 41 3.2.1 The Impact of Surface Roughness ................................ ................. 51 3.3 Discussion ................................ ................................ ................................ .. 57 Chapter. 4: Protein Adsorption to Surfaces ................................ ................................ ...... 58 4.1 The Aggregation and Activation of Platelets ................................ ............. 58 4.2 The QCM Technology ................................ ................................ ............... 60 4.3 Protein Adsorption Assessment Using QCM D ................................ ........ 62 4.4 Results ................................ ................................ ................................ ........ 65


ii 4.5 Discussion ................................ ................................ ................................ .. 68 Chapter. 5: Microvascular Endothelial Cell Proliferation on Semiconductor Substrates ................................ ................................ ................................ ...... 70 5.1 Microvascular Endothelial Cells and the Vessel Internal Lumen .............. 70 5.2 Cell Proliferation in Vitro : Experimental Protocol ................................ .... 73 5.3 MTT Assay and Fluorescent Microscopy Results ................................ ..... 78 5.4 Discussion ................................ ................................ ................................ .. 82 Chapter. 6: Conclusions and Future Work ................................ ................................ ....... 84 6.1 Conclusions: 3C SiC Versus Si ................................ ................................ 84 6.2 Future Work ................................ ................................ ............................... 87 References ................................ ................................ ................................ .......................... 90 About the Author ................................ ................................ ................................ ... End Page


iii List of Tables Table 1.1. Electrical, thermal and physical properties of SiC and Si [18] ........................ 10 Table 2.1. A sample of primary in vitro tests, Adapted from [51] ................................ ... 26 Table 3.1. Summary of samples roughness expressed as [average roughness() + (nm) ................................ .. 51


iv List of Figures Figure 1.1. (Top) Flow diagram of the clotting cascade showing both, extrinsic and intrinsic pathways [6].. ................................ ................................ ............. 5 Figure 1.2. Platelet lamellipodia and filopodia after activation on a surface. Adapted fr om [6] ................................ ................................ ............................. 7 Figure 1.3. Sequence of events that leads to formation of fibrous capsules around sensors [16]. ................................ ................................ ........................ 9 Figure 1.4. Fluorescent micrographs of (a) B16 melanoma cells (b) BJ fibroblast cells and (c) HaCaT keratinocyte cells on SiC substrates 72 hours after s eeding [9]. ................................ ................................ ........................... 13 Figure 1.5. Fluorescent micrographs of (a) B16 melanoma cells (b) BJ fibroblast cells and (c) HaCaT keratinocyte cells on Si substrates 72 hours after seeding [9]. ................................ ................................ ................................ .... 13 Figure 1.6. Histogram of cell proliferation of B16, BJ and HaCaT cells, norma lized to the control sample, expressed as x MTT assays at the third day. ................................ ................................ ......... 14 Figure 1.7. Histogram representing re sults of neuronal H4 and PC12 cell proliferation tests. ................................ ................................ ......................... 17 Figure 1.8. 45x45 um AFM micrographs of H4 cells deposited on different substrates. ................................ ................................ ................................ ...... 19 Figure 2.1. Scanning electron micrographs showing platelet morphology of Si (left) and SiO 2 (right). ................................ ................................ ................... 30 Figure 2.2. Number of adhered platelets per 15 mm 2 for Si, Si based materials, SU 8, parylene and polyurethane. ................................ ................................ 31 Figure 2.3. Fibrinogen promoting platelet platelet adhesion by binding glycoprotein IIb IIIa (GPIIb IIIa). ................................ ................................ 32 Figure 2.4. SEM micrographs showing endothelial cells on three types of Cr DLC. Cr is 1,5 and 10% [62] ................................ ................................ ........ 34 Figure 3.1. Diagram of a red blood cell (erythrocyte) shape, ................................ ......... 36


v Figure 3.2. Classification of the 5 types of white cells [75]. ................................ ........... 36 Figure 3.3. S canning electron microscope images of resting (left), partially (center) and fully activated (right) human platelets. ................................ ..... 38 Figure 3.4. Diagram of a tube of anticoagulated blood that has settled, illustrating the separation of blood into three layers. ................................ .... 39 Figure 3.5. Preparation of platelet concentrates using the PRP and BC techniques [15]. ................................ ................................ ............................. 40 Figure 3.6. Illustration of the image analysis process for platelets seeded on a Si sample. ................................ ................................ ................................ .......... 44 Figure 3.7. Fluorescent micrographs comparing adhered platelets on 3 different SiC polytypes and Si. ................................ ................................ .................... 45 Figure 3.8. Fluorescent micrographs revealing activated platelet clumps (left), non clumped activated platelets (center) and a large non activated platelet clump on the surface of 4H SiC samples. ................................ ........ 46 Figure 3.9. Statistical analysis of particle circularity on 3 different SiC polytypes and Si.. ................................ ................................ ................................ .......... 49 Figure 3.10. Statistical analysis of particle counts on 3 different SiC polytypes and Si. ................................ ................................ ................................ ................. 49 Figure 3.11. Statistical analysis of particle size on 3 different SiC polytypes and Si.. ................................ ................................ ................................ ................. 50 Figure 3.12. Fluorescent micrographs comparing Si (left) and 3C SiC (right) samples revealing higher degree of thrombogenicity in Si than in 3C SiC. ................................ ................................ ................................ ................ 50 Figure 3.13. AFM micrograph taken in tapping mode. ................................ ..................... 52 Figure 3.14. AFM micrograph taken in tapping mode. ................................ ..................... 52 Figure 3.15. AFM micrograph taken in tapping mode. ................................ ..................... 53 Figure 3.16. AFM micrograph taken in tapping mode. ................................ ..................... 53 Figure 3.17. AFM micrograph taken in tapping mode. ................................ ..................... 54 Figure 3.18. Fluorescent micrograph comparison of two low roughness surfaces. .......... 55 Figure 3.19. Fluorescent microgr aph comparison of two high roughness polycrystalline 3C SiC surfaces. ................................ ................................ ... 56


vi Figure 3.20. Fluorescent micrograph of platelet adhesion on a low roughness silicon(100) surface ................................ ................................ ....................... 56 Figure 4.1. Platelet aggregation mediated by the action of fibrinogen and von Willebran d Factor (vWF) [87] ................................ ................................ ...... 59 Figure 4.2. Diagram of a quartz crystal sensor [91] ................................ ........................ 60 Figure 4.3. Sequence of steps of a typical hemocompatibility assesment using a Q sense device. ................................ ................................ ............................ 65 Figure 4.4. Histogram of the fribrinogen and GPIIb IIIa adsorption to gold, silicon and 4H SiC substrates. ................................ ................................ ...... 66 Figure 4.5. ANOVA tables of the statistical data for Fibrinogen (top) and GPIIb IIIa (bottom).. ................................ ................................ ................................ 67 Figure 4.6. Histogram of the fribrinogen and GPIIb IIIa mass (ng/cm 2 ) adsorption to gold, silicon and 4H SiC substrates. ................................ ....... 68 Figure 5.1. Fluorescent micrographs illustrating the pavement like pattern of the endothelial cells. ................................ ................................ ........................... 70 Figure 5.2. Diagram of a small artery in cross section. ................................ ................... 72 Figure 5.3 Schematic representation of s ample position for cell plating experiments. ................................ ................................ ................................ .. 76 Figure 5.4. (Top) Histogram displaying th e normalized results (respect to the treated wells) of the MTT assay after 72 hours. ................................ ........... 78 Figure 5.5. Fluorescent micrographs of HMVE C proliferation on various substrates (treated well, non treated well, 3C SiC and Si) after 72 hours. ................................ ................................ ................................ ............. 79 Figure 5.6. (Top) Histogram displaying the normalized results (respect to the treated wells) of the MTT assay after 120 hours. ................................ ......... 79 Figure 5.7. Fluores cent micrographs of HMVEC proliferation on various substrates (treated well, non treated well, 3C SiC and Si) after 120 hours. ................................ ................................ ................................ ............. 80 Fi gure 5.8. (Top) Histogram displaying the normalized results (respect to the treated wells) of the MTT assay after 168 hours. ................................ ......... 81 Figure 5.9. Fluorescent micrographs of HMVEC proliferation on various substrates (treated well, non treated well, 3C SiC and Si) after 168 hours. ................................ ................................ ................................ ............. 81


vii Hemo compatibility Assessment of 3C SiC for Cardiovascular Applications Norelli Schettini ABSTRACT The hemocompatibility of crystalline Silicon Carbide (SiC), in its cubic form ( i.e., 3C SiC) has been evaluated and compared to Silicon (Si), the leading material in biosensing applications. Silicon carbide (SiC) is a hard, chemically robust material, very well suited for harsh environment applications, and has been suggested to have very good bi ocompatibility. Additionally SiC in its amorphous form, has been used as a coating for medical implantable devices such as bone prosthetics and cardiovascular stents. However, assessment of single crystal 3C SiC for cardiovascular applications has not bee n reported. In this research we have studied the interactions of single crystal 3C SiC with platelets and human microvascu lar endothelial cell (HMVEC) to assess the degree of hemocompatibility of 3C SiC The more hemocompatible a material is, the less pla telet adhesion would be expected. Using fluorescence microscopy higher platelet adhesion was statistically observed on Si than on SiC. In addition 3C SiC surfaces showed less platelet reactivity, measured by the degree of platelet adhesion, aggregation and activation with mostly circular morphology of adhered platelets while Si showed an elevated presence of non activated (Circular) platelet clumps.


viii Additionally, HMVEC proliferation assessment suggest that 3C SiC performs comparably to high attachment cul ture wells with enhanced proliferation without affecting cell morphology These results suggest that 3C SiC is a promising candidate for applications in the blood stream due to its low thrombogenic characteristics and good hemo compatibility


1 Chapter. 1: Introduction Biocompatibility is the required property of materials and systems that are intended to be in contact with biological matter to perform according to the system design but without interfering with the integrity of th e t issues and biological matter that such material is in contact with and without generating any adverse reactions [ 1 ] Silicon (Si) has been long used in the construction of Microelectromechanical systems, MEMS devices. S ubsequently; the question of whether silicon is or is not a biocompatible/suitable material for biotechnology applications or implantable devices is a crucial question in the area of microfabrication and biomedical sensing devices. Although this question seems a simple one, and the kind that one should be able to solve with some literature search, this search usually tends to provide answers as complex as the biological systems that we try to work with. First, on e cannot talk about a biocomp atible material in an absolute way as a material that performs extra ordinarily well for some biological matter can be devastating for others and additionally there is a lot of uncertainty about the mechanisms that determine biocompatibility as there as ma ny ways in which tissues and materials interact [ 2 ] I n the case of silicon previous research shows that the material can perform well for certain tissues but can be cytotoxic for others [ 1 ] [ 3 ] [ 4 ] Many limitations have been reported for devices fabricated with Si, including carcinogenic effects due to the presence of minute silicate particulates, inflammatory responses and device encapsulation. The limitations of how compatible a material is when in contact with tissues is limited by the application e.g., issue engineering, implantable sensors, drug del iv ery, etc. I f the


2 application to be developed requires the proliferation of cells on the surface of a device, such as grafts, cell viability can be considered a pretty good measurement of a material's biocompatibility, but if the material is going to be in contact with blood then we need measurements of the degree of thrombogenicity defined as the tendency of a material in contact with blood to produce or induce thrombus formation [ 1 ] [ 5 ] Biocompatibility is also defined in terms of behaviors that are desirable for a material in contact with biological matter, e.g., a material being non toxic, non immunogenic, non carcinogenic, non irritant and so on Another factor of supreme importance is the implant life time. In the case of short term implants advanced biopolymers offer excellent biocompatible surfaces, but this solution is not suitab le for long term implants as the s e biopolymers tend to degrade in the hars h environment of the human body. In addition, polymers also lack the sensing capabilities that semiconductor materials have, this is especially important in the case on bio sensing implantable devices. The particular case of material biocompatibility with blood (and the cardiovascular system in general ) is called hemocompatibility and this research explores different aspects of the biocompatibility of single crystal 3C SiC as a potential material for future cardiovascular appl ications due to its less bioreactive nature when in contact with biological samples and its high degree of proven chemical resilience. 1.1 Research Objective and M otivation Heart failure and stroke are two of the leading causes of death in the United States every year. Today, patients with cardiovascular conditions can benefit from new medication s improved means of cardiovascular monitoring and treatments which were unavailable in the past; however there is still room for improved cardiovascular


3 therapeutic devices based on biotechnology, ranging from real time detection and monitoring of cardiac event markers to accurately timed drug deliv ery and dosage. Unfortunately, current technology does not allow for their realization due to the insufficient performanc e of the present biomaterials Clearly, there is a critical need for a new class of biomaterials that are hemocompatible and that could also allow the fabrication of MEMS based sensing devices Developing a sensor that operates in real time and in vivo rep resents a challenge dominated by the hemocompatibility of the materials used and the successful accomplishment of different r equirements such a s matching blood flow dynamics, duration of blood contact, selection of an appropriate sensor geometry, and senso r surface quality (in terms of surface roughness) and chemistry (i.e., specific adhesion of the target molecule). The formerly mentioned requirements are necessary to avoid device induced thrombus formation which can be life threatening either because of d evice failure or because of thrombi embolization [ 6 ] [ 7 ] Silicon carbide (SiC) is a hard, chemically robust material, very well suited for harsh environment applications such as implantable devices in the human body [ 8 ] [ 9 ] Recent assessment of the single crystal form o f this material (by Dr. C. Coletti and Dr. C Frewin in the SiC and Drug and Gene Delivery laboratories at USF ) suggests that it has very good biocompatibility [ 9 ] [ 10 ] [ 11 ] Moreover, SiC, in its amorphous form, has been used previously as a coating for medical implantable devices such as bone prosthetics and cardiovascular stents [ 8 ] [ 12 ] H owever, assessment of single crystal SiC for cardiovascular applications has not been reported and is the focus of this dissertation research


4 Single crystalline SiC grows in ma ny crystal forms, but only one form of SiC is both cubic and can be grown on single crystal Si substrates 3C SiC. Single crystal 3C SiC is highly resistant to harsh environments and it is chemically inert I t can be grown heteroepitaxially on silicon s ubstrates and standard silicon processing can be used to micro machine this material [ 13 ] However in order to exploit these material properties for a blood stream implantable device, its biocompatibility with bloo d and vascular endothelial lining must be assessed. This research looks for establishing whether or not single crystal 3C SiC is a hemocompatible material suitable for cardiovascular app lications, through the explorations of its platel et adhesion and biocompatibility with human microvascular endothelial cells. Next, an introduction to thrombus formation, biocompatibility testing and the biocompatibility of 3C SiC is presented to further understand the processes behind thrombogenicity and the rationale behind the different aspects that we have chosen to evaluate to determine the degree of hemocompatibility of 3C SiC. 1.1.1 Thrombus F ormation Device hemo in compatibility occurs when a blood contacting man made device induces thrombus formation [ 6 ] Almost immediately after injury or the introduction of a foreign material into the body clottin g factors become activated (see figure 1. Top) and induce the generation of an enzyme called thrombin [ 14 ] Thrombin converts the soluble protein fibrinogen into insoluble fibrin. The formed fibrin provid es the scaffold for the entrapment of platelets, blood cells, and plasma proteins (indeed this is critical for patient survival as it is the mechanism behind blood clotting).


5 In the case of a blood contacting device, key blood proteins, such as fibrinogen and adsorb onto the surface of the device (Figure 1. Bottom), This is followed by the adhesion of platelets to these proteins on the surface of the implanted probe, through ke y receptors on the activation; such binding to surface adsorbed proteins plays an important role in the clotting cascade [ 15 ] [ 16 ] [ 17 ] Figure 1 1 (Top) Flow diagram of the clotting cascade showing both, extrinsic and intrinsic pathways [ 6 ] (Bottom) Sequence of events that leads to throm bus formation on the surface of a man made material placed within t he blood stream [ 16 ]


6 subsequently, that the initiation signal for platelet deposition and activation is e xposure of the underlying vessel wall, that are normally hidden from circulating platelets by a non compromised endothelial lining [ 15 ] Additional parameters determining platelet response are the depth of the injury, the vascular bed, age o f the individual, hematocrit value (blood volume occupied by red cells) the speed of blood flow and the size of the blood vessel [ 15 ] For platelet plug formation to occur, platelets must undergo adhesion, activation and aggregation [ 18 ] [ 19 ] Most platelet activators are released or synthesized at the place of the vascular injury. Additionally, biochemical interactions between platelets and erythrocytes (red cells) also enhance platelet activation. Adhesion of platelets to subendothelial stru ctures may lead itself to platelet activation [ 6 ] [ 15 ] Once activated, platelets suffer a change in their discoid shape which occurs in response to ma ny different agents, resulting in a transformation to a spiny sphere with long, thin filopodia extending several micrometers out from the membrane surface [ 15 ] When platelets adhere to surfaces, they gen erally undergo spreading and activation which depends on the protein surface on which they spread to, with collagen inducing the most activation. Platelet spreading results in the development of broad lamellipodia rather than spike like filopodia (see Figu re 2) which is accompanied by granule secretion [ 15 ] Intravascular placement is the most desirable for most chemical measurement. However, p latelet role in the vascular tree healing process, as well as the processes that intervene in platelet adhesion, spread ing, aggregation and activation, all must be considered when studying the issue of device implantation In addition there is the role


7 played b y platelet activators and glycoprotein ligands, platelets changes and surface proteomics. These are complicated bio physical processes that must be well understood before considering the placement of a man made sensor or material into the cardiovascular system. The biological response of the body to foreign mate rials determines the success of implantable biosensors and the degree of their bio/hemocompatibility. Once the blood clot is formed, the clot itself promotes more clotting; the platelet thrombi appear capable of rapidly recruiting tissue factor (thrombopla stin) from the blood which is associated with P selectin glycoprotein ligand (PSGL 1), a ligand for the P selecting expressed on the surface of activated platelets. Activated platelets also recruit neutrophils and monocytes expressing PSGL 1 [ 15 ] Figure 1 2 Platelet lamellipodia and filopodia after activation on a surface Adapted from [ 6 ] Activation of the clotting cascade by intravascular sensors presents risks to the patient in that emboli may be shed and clog blood vessels elsewhere in the body [ 6 ] Therefore, careful selection of suitable hemocompatible material is the first step that must be undertaken before developing any sensor for the cardiovascular system.


8 From the sensor device data collection perspective, the presence of activated platelets on the surface of an implanted ch emical sensor can create a local chemical environment that differs from that of bulk blood. Even in the absence of complete thrombus formation on the surface of the device, a monolayer of active cells can cause significant measurement errors; hence, the ac tivity of cells adhered to the surface of any implanted sensor can act like an immobilized enzyme catalyst to create a similar local change in key analyte levels [ 16 ] The former implies the need for a m aterial for sensor device fabrication that not only reduces thrombus formation (or eliminates it in the best case scenario) ; by controlling protein activity on the surface of the device, but also that minimizes platelet adhesion. The use of an immobilized heparin coating on the surface of the implanted device, which is the current str ategy used by most researchers [ 20 ] is not a solution either, because such coatings have been shown to not reduce platelet activation and adhesion on foreign materials. Heparin coating s as an anticoagulant does reduce the size of the thrombus formed on the device by locally inhibiting the activities of thrombin and Factor Xa, however it will not significantly diminish the pres ence of adhered, metabolically active cells [ 16 ] and, moreover, it will be consumed over time, so it is clearly not a good option for long term implantable devices. Implantable sensors are subject to an a cute inflammatory response that involves the release of fluid and plasma proteins followed by recruitment of leukocytes to the implant site.


9 Figure 1 3 Sequence of events that leads to formation of fibrous capsules around sensors [ 16 ] Leukocytes (mainly neutrophils and monocytes) migrate to the implanted device and adhere to the surface in an attempt to begin the process of phagocytosis. Finally, a fibrous capsule composed mainly of macrophages and collagen forms around the implanted device. The capsule can alter local analyte levels via metabolic reactions [ 16 ] Different approaches can be used to determine and modify the degree of thrombogenicity of MEMS materials, including biomolecular surface interactions. For vascular implants, human microvascular endothelial cells and platelets must be used to assess the hemocompatibility of any foreign material inserted into the vascular system. Endothelial cells relate directly to platelet interactions and bio/hemocompatibility of surface and does not promote platelet attachment. The interactions of both end othelial cells and platelets with a desirable material for device implantation must be studied to get with these foreign surfaces [ 21 ]


10 1.2 SiC Properties SiC is a wide bandgap (great sensing potential), optically transparent semiconducting material that exhibits great electrical properties in harsh environments, such as high temperature elasticity and chemical inertne ss [ 22 ] Table I displays some of the various properties of silicon carbide. Its electrical, mechanical and chemical properties (i.e., resistance to corrosion and wear, etc.) make it an ideal candidate for sensing applications, such as MEMS devices. Table 1 1 Electrical, thermal and physical properties of SiC and Si [18] SiC forms in three main crystal orientations; cubic (3C SiC), hexagonal (4H SiC and 6H SiC), and rhombohedral (15R SiC) where 4H 6H and 3C SiC are the main polytypes of practical and commercial interest [ 22 ] As SiC is not a chemical compound normally found in nature it must be artificially generated. One of the methods used for the creation of single crystal, high quality SiC material is provided through the use of a 3C SiC 4H SiC 6H SiC Si Bandgap (eV) 2.36 3.23 3.0 1.12 Breakdown field (MV cm 1 ) 1 3 5 3 5 0.25 Electron Mobility (cm 2 V 1 s 1 ) at 300K <800 <900 <400 1500 Hole Mobility (cm 2 V 1 s 1 ) at 300K <320 <120 <90 450 Thermal conductivity (W/cm K) 5 4.9 4.9 1.5 Thermal expansion @300 K @1500 1600 K (10 6 K 1 ) 3.8 5.5 NA 4.3 c axis c axis 2.6 4.56 4.6 300 to 700 47


11 chemical vapor deposition (CVD) on a bulk crystal which is performed at high temperatures [ 23 ] 3C SiC is the only single crystal form of SiC that can be grown heteroepitaxially using Si as a substrate, due to the similarities in the crystal lattice structures (i.e., crystal lattice). In addition 3C SiC is already a well established semiconductor and MEMS material [ 22 ] [ 24 ] 3C SiC also has the benefit of being compatible with most Si micromachining techniques, such as reactive ion etching (RIE) and ion implantation [ 24 ] [ 25 ] T o date, few crystalline semiconductor materials have been assessed for their biocompatibility. Si, the leading material in the MEMS field, has been shown to be cytotoxic, due to the formation of silica and silicates in aqueous solutions [ 3 ] [ 26 ] [ 27 ] [ 28 ] [ 29 ] Other materials, which do not undergo hemo in compatibility problems but that are not suitable for sensing applications, have been used as coatings for short term contact applications. Materials such as polymers are consumed over time when implanted in the body, so the y do not offer an alternative for long term implantable application s even though they are the most developed material for this application. One advantage of such materials is their very low cost, which is import ant for short time, high volume (device quantity) implantable applications. There is evidently a need for a crystalline material with suitable characteristics for sensing applications in harsh environments, such as in bodily fluids, which also provides a b io/hemocompatible surface.


12 SiC, in its amorphous form, has been used successfully as a coating for prosthetic devices, it is one of the most used materials for cardiovascular applications, and has been shown to be highly effective in vivo clinical trials as a coating for heart stents [ 8 ] [ 12 ] [ 21 ] [ 26 ] [ 30 ] [ 31 ] [ 32 ] [ 33 ] The literature also reveals that collagen deposition on SiC whiskers is non progressive and that it generates an inflammatory response that decreases gradually after implantation [ 32 ] All of the above s uggest attractive bio/hemocompatibility properties of this material in its amorphous form and one could suspect that 3C SiC, in its single crystal form should evidence the same properties; however, before the research conducted by Dr. C. Coletti of the Si C group at the University of South Florida, no previous research on the biocompatibility of single crystal 3C SiC had been reported in the literature [ 9 ] [ 10 ] [ 34 ] 1.2.1 SiC a Promising M aterial for B iosensing A pplications in the B lood S tream SiC is a polymorphic material, having approximately 170 different crystal organizations, but sharing a common stoichiometry (i.e., all forms contain bi layers of Si and C) [ 22 ] Two of the hexagonal polyt ypes, 4H SiC and 6H SiC have been fashioned in the past into electrode myocardial sensors and have shown low biofouling, non toxicity and in general good biocompatibility [ 35 ] [ 36 ] Amorphous and polycrystalline forms of silicon carbide, a SiC and poly SiC, have been used as coatings for many implantable biomedical devices, and clinical studies for bone prosthetics and heart stents have confirmed its bio compatibility. Dr. C. Coletti researched the in vitro biocompatibility of three lines of cells, namely B16 F10 mouse melanoma cells (ATCC CRL6475), BJ human fibroblasts (ATCC CRL2522) and human keratinocytes (HaCaT) by plating and culturing them in the


13 ap propriate media onto the surfaces of the main polytypes of SiC: 3C SiC, 4H SiC, and 6H SiC, and assessing cell morphology, attachment to the substrate surface, successful cellular binding and proliferation [ 9 ] In her research, Coletti used Si and standard culture plates, as negative and positive controls respectively, which served as benchmarks for biocompatibility comparisons [ 9 ] These three different cells lines were selected for their different adhesion properties, i.e., B16 melanoma cells are very aggressive, proliferate at a fast rate and adhere indifferently on substrates with different degrees of biocompatibility, conversel y, BJ fibroblasts are much more selective as to a preferred substrate for adherence and should provide more [ 9 ] Figure 1 4 Fluorescent micrographs of (a) B16 melanoma cells (b) BJ fibroblast cells and (c) HaCaT keratinocyte cells on SiC substrates 72 hours after seeding [ 9 ] Figure 1 5 Fluorescent micrographs of (a) B16 melanoma cells (b) BJ fibroblast cells and (c) HaCaT keratinocyte cells on Si substrates 72 hours after seeding [ 9 ]


14 Figure 1 6 Histogram of c ell proliferatio n of B16, BJ and HaCaT cells, normalized to the control sample, expressed as x measured via MTT assays at the third day. Note that cell proliferation is greater on SiC than on Si [ 9 ] using separate MTT [3 (4,5 Dimethylthiazol 2 Yl) 2,5 Diphenyltetrazolium Bromide] assays for the three different cell lines The adherent cell morphology was studied using fluorescence microscopy. SiC samples, Si samples, and cell culture wells were plated and t he cell morphology was inspected after 4, 24, 48, 72 hours for the B16 cells and 4, 24, 48, 72, 96 and 120 for the BJ and HaCaT cells. The cells were fluorescently labeled with CMFDA (5 chloromethylfluorescein diacetate) cell tracker dye (Molecular Probe s, Eugene, OR) [ 9 ] Optical inspection revealed that both the B16 and BJ cells were well distributed on the SiC surfaces after 4 hours of incubation, elongated and flattened against the surface after 24 hou rs, and became confluent after 3 days (B16) and 5 days (BJ and HaCaT) independently of polytype used, and chemical composition [ 9 ] Images of B16, BJ and HaCaT cells cultured on SiC (72 hours after seeding), where the adherent cells appear to


15 be flattened and stellate shaped, maximizing the contact area with the substrate and similar to that observed in culture wells, are shown in Fi gure 4 (a, b and c). The lack of SiC cytotoxicity was confirmed by the fact that the cells could be cultured to confluence on all the SiC substrates used [ 9 ] The morphology and density of B16, BJ and HaCaT cells cultured on Si substrates under identical conditions revealed that B16 cells tended to form in clumps, while the adherent cell quantity, in the case of BJ fibroblasts and HaCaT keratinocytes was drastically reduced (see Figure 5). In addition, the reduced contact area of the adherent cells (low adhesion force) suggests some degree of cytotoxicity for the Si substrates [ 9 ] Figure 6 shows a histogram of the calculated cell viability from MTT assays, performed on the third day for the three different cell lines, which confirmed the optical inspection results. The histogram clearly shows that SiC is a high quality surface for cell culture with significantly better performance than Si. No statistically s ignificant differences in cell proliferation were found among the different SiC polytypes [ 9 ] An extremely satisfying proliferation was observed for BJ cells on SiC [ 9 ] From the histogram it is also clear that the B16 cell proliferation is lower than the BJ cell proliferation on SiC substrates but always higher than on Si. As in this study the variation of roughness of the different substrates was very sma ll compared to the cell dimensions Therefore future work regarding surface roughness effects on the biocompatibility must be performed, with roughness values exceeding the cell diameter. Surface roughness also plays a role in hemocompatibility, as high surface roughness can lead to blood cells lysis and induce clotting.


16 After the results of Dr. Coletti the biocompatibility of 3C SiC seemed promising. Subsequently, Dr. Frewin, also in the USF SiC group, engaged in the research of the neuron SiC interface and the development of Brain Machine Interface (BMI) devices. In his research Dr. Frewin proposed cubic silicon carbide (3C SiC) as a solution to the biocompatibility problems of current implantable BMI devices. Dr. Frewin's study [ 11 ] showed that both materials possess good biocompatibility with neural cell lines (human neuroglioma and PC12 rat pheochromocytoma cell lines) that, in the worst case, is comparable to cell treated polystyrene and, in most cas es, even surpasses polystyrene. Additionally, in this research the action potential of differentiated PC12 cells was activated using 3C SiC within an electrode device. H4 human neuroglioma and PC12 rat pheochromocytoma cell lines were used for the study, a nd polystyrene (PSt) and amorphous glass were used as controls or for morphological comparison. MTT assays were performed to determine general cell viability with each substrate and atomic force microscopy (AFM) was used to quantify the general cell morpho logy on the substrate surface along with the substrate permissiveness to lamellipodia extension. 3C SiC was the only substrate tested to have good viability and superior lamellipodia permissiveness with both cell lines, while nano crystalline diamond ( NCD ) showed a good level of viability with the neural H4 line but a poor viability with the PC12 line and lower permissiveness than 3C SiC.


17 Figure 1 7 Histogram representing results of neuronal H4 and PC12 cell proliferation tests. The graph is expressed as the s.d.m., with error bars indicating the s.e.m., from the three MTT assays. From [ 11 ] The MTT analysis results for both the H4 and PC12 cell lines are displayed in Figure 1.7. The H4 cell line shows a viab ility level of approximately 80% + 5% as compared to the PSt control for all samples excluding S i (111) which performed at 108% + 9%. Si (10 0) 3C SiC surfaces and NCD sample means were found to be statistically similar, as well as Si ( 111) and PSt. Statistically significant similarities were found between 3C SiC samples and Si (111) when testing for PC12 proliferation, showing proliferation values of approximately 90% of the PSt control, while the NCD and Si (100) samples were found to be statistically similar with proliferations of 46 and 55%, respectively. The morphology and lamellipodia attachment quality were evalu ated with AFM analysis Figure 1.8 displays two selected AFM micrograph scans of the H4 cell line incubated on each of the tested substrates with one scan showing a living cell and the other a fixed cell. The H4 cells on PSt show a low level of attachment to the surface Figure 1.8(a) indicates that H4 cells on PSt have microtubule extensions and the average


18 lamellipodia extension over the whole cell group is 28.9% of the cell area while alive. Un fortunately, as seen in Figure 1.8 ( b ) the lamellipodia coverage is reduced by 13.5% after cell fixation for. The H4 cells on glass, as seen in Figure 1.8 ( c ) for the living cell and Figure 1.8 ( d ) for the fixed cell, react similarly to the cells on PSt. Although there are large expanses of lamellipodia present with many of the cel ls on the glass substrate, they show very little attachment to the surface, constituting an average of 45.5% of the cell in the living cell scans, and only 8.43% of the cell after fixation (a 27% reduction). The H4 cells on the surface of Si (Figure 1.8 ( e ) and 1.8 ( f ) ) on 3C SiC (Figure 1.8 ( g ) and 1.8 ( h ) ), and on NCD (Figure 1.8 ( i ) and 1.8 ( j ) ) possess similar levels of cell attachment to the surface, but they show some differences in the lamellipodia attachment. The H4 cells flatten and expand over a larger surface area on these substrates. The H4 cell s on Si show the lamellipodia extending from many parts of the cell accounting for an average of 36.4% of the cell area when measured alive and 26.4% after fixation. NCD shows a slightly better lamellipodia ext ension and attachment level with 30.5% of the cell area before fixation and 22.4% cell area after fixation.


19 Figure 1 8 45 x 45 um AFM micrographs of H4 cells deposited on different substrates Each micrograph possesses a pseudo three dimensional insert in the upper right hand corner to give some perspective on cell morphology. (a), (c), (e), (g), and (i) are H4 living cells measured in PBS deposited on polystyrene, glass, Si, 3C SiC, and NCD. (b), (d ), ( f) (h), and (j) are H4 cells that have been fixed with 2% formaldehyde on polystyrene, glass, Si, 3C SiC, and NCD. From [ 11 ] 3C SiC has the smallest measured surface area for the lamellipodia, which measured 15.1% of the living cellular area, but was the only sample group that increased in area percentage and was 15.4% after the cell was fixed. 3C SiC cells also show multiple la rge


20 projections of microtubule extensions on almost all of the cells measured. The lamellipodia reduction level for Si and NCD was 10 and 8.1%, respectively, while 3C SiC shows a mean increase of 0.3%. The PC12 cells have much different attachment charact eristics than the H4 cells and this is most likely due to the fact that they were not changed into peripheral neurons through the addition of neural growth factor (NGF). Glass, Si, and NCD all show lower attachment characteristics for this cell line. Analy sis of the cell height to area ratios and attachment angles for glass, Si, and NCD, indicat ed that the cells are trying to minimize their contact with these substrates which was confirmed as each of these substrates show poor lamellipodia attachment. Altho ugh Si shows good extensions of microtubules, it has a lamellipodia reduction of 10.2% area as indicated for the 25.0 and 14.8% live cell and fixed cell lamellipodia measurements. The PC12 cells on 3C SiC have a slightly lower height to surface area ratio and a lower angle of attachment when compared to the previous three samples. The lamellipodia shows a much better attachment factor composing 29.3% of the cell area for the living cell and 24.8% of the cell area for the fixed cells which amounts to only a reduction of 4.5%. Cells on 3C SiC also possess multiple large microtubule extensions. The PC12 cells showed a good permissiveness with PSt, t his material allows the cell to project multiple microtubule extensions and had a 3.3% increase in lamellipodia a ttachment area due to the living cell area lamellipodia percentages were 25.9 and 29.2% for the living and fixed cells, respectively.


21 Frewin's research also showed that after cell seeding and proliferation on the substrates, the Si surface suffer ed physica l damage that was visible through AFM analysis of the samples, this was not the case for 3C SiC 1.3 Summary and D issertation O rganization The demo n strated biocompatibility of 3C SiC, together with its other properties, such as mechanical and chemical resilience, the capability to built sensing devices or patterned grafts using traditional microfabrication techniques make this material an ideal candidate fo r cardiovascular application s However, such applications cannot be realized until it is known for certain if th is material stimulates thrombogenic behavior that would induce thrombus formation. This research looks to assess the hemocompatibility of this material so that its use in cardiovascular applications can be pursued The subsequent chapters explore the nature of blood, its formed elements (§3.1) and the formation of c lot s and how it relates to interactions with foreign surfaces (§2.1) ; additionally we introduce the different techniques and rationales behind hemocompatibility testing (§2.2 and §2.3) and finally the results of this research by explori ng three different perspectives: platelet adhesion (§3) protein adsorption (§4) and interactions with microvascular endothelial cells (§5)


22 Chapter. 2: Biocompatibility of M aterials A material, synthetic or natural, that is used in contact with a biological system and that is ab le to co exist without causing an unacceptable degree of harm to the system is called a biomaterial [ 16 ] E arly biomaterials included silicone, polyurethanes, rubber, Teflon nylon, titanium, and stainless steel These early materials were used to develop various types of prosthetic devices such as biodegradable scaffolds, cochlear implants, breast implants, shoulder, hip and knee prostheses, dental and intraocular implants, etc. Biomaterials have improved the quality of life of many patients around the world and have prov ided alternatives to the limitations of tissue rejection or limited organ donors (for the cases in which it applies). The development of biomaterials has been full of success but a lso of challenges related to material rejection or incompatibility Most mammalian cells require attachm ent to an external surface or they will experience apoptosis. The interaction of the cells and the material at the interface mu st induce cell attachment and proliferation without inducing negative morphological changes or intoxicating the cell. Although the interaction between cells and substrates is complex and it is regulated by the interactions of the material with the extrace llular matrix, cell membrane, cell cytoskeleton etc., it is widely accepted that a cell in contact with a biocompatible material will attach to and spread out on the surface and then proliferate thus making cell morphology and viability good measurements of biocompatibility [ 11 ] [ 37 ] [ 38 ] C ells on a non permissive surface will interact with the substrate by decreasing their attachment


23 area due to flow back of actin filaments from the cytoskeleton into the central core of the cell [ 37 ] [ 39 ] Unfortunately, the above concepts of viability and morphology as a measurement of biocompatibility cannot be extended to the interaction of materials and blood. If a blood conta cting material allows blood organelles to attach and s pread on its surface thi s would alter the normal blood flow create turbulence and ultimately induce thrombus formation Thrombus formation has l ife threatening consequences for the patient due to thrombus embolization and must be avoided at all cost Therefore a new set of concepts to describe the biocompatibility of a substrate with blood must be utilized. The following sections of this chapter describe these concepts and the different perspectives on hemocompatibility evaluation. 2.1 Biocompatibility of Materials with B lood: Hemocompatibility Hemocompatibility concerns to the properties of materials suitable for blood contacting applications. Blood contacting devices range from needles to vascular grafts, cardiopulmonary bypass and heart valves among others. Highly invasive devices usually extend the life of patients and perform reasonably well for periods of time before clinical complications can be observed These devices also require post implant clinical treatments like the use of blood antico agulants Today's blood compatibility philosophy is to guarantee the safety of the implant and to minimize post implant procedures which finally represent long term benefit for the patient. [ 1 ] [ 40 ] [ 41 ] The most common device/material failures due to a lack of hemocompatibility are related to thromboembolic complications Other complications of prosthetic devices in contact with the blood stream include erosion of the material, mechanical failure and


24 nons tructural dysfunction hemolysis and infection. H owever for the purpose of studying hemocompatibility it is important to fo cus on reactions that are specific to the placement of a material in contact with blood and not to those that may be the result the activation of the multiple defense mechanisms that the body has available in response to the presence of foreign devices [ 42 ] [ 43 ] [ 44 ] These specific reactions are called thrombotic and thromboembolic complications. To study hemocompatibility is related to the study of the thrombogenic behavior of a material; a material with low thrombogenicity will be therefor e more hemocompatible and will also reduce the risk of bleeding associated with the use of blood thinners [ 45 ] [ 46 ] [ 40 ] To date a man made material that is purely hemocompatible has not been reported and the mechanisms behind surface blood interactions have not been completely understood. Indeed, it may be this later point that has made the dev elopment of a truly hemocompatible material so difficult. Smooth polymeric surfaces have been used as a coating for materials used in blood pumps and they have shown that roughness seems to play an important role in hemocompatibility Since the introduction of these polymeric coatings based o n the use of anticoagulants like heparin or thrombin inhibitors [ 47 ] [ 48 ] two approaches have been use d to address the subject of surface blood in teractions. First, t he creation of a pseudointima formed from the blood lining and, second, t he reduction of blood formed elements and protein interactions with the surface. A material that has a high compatibili ty with endothelial cells and that possess es low thrombogenicity, ultimately expressed in fewer platelet adsorption, aggregation and activation would satisfy both


25 approaches and would be desira ble for the development of long term cardiovascular devices [ 7 ] [ 46 ] The ISO 10993 4 standard introduce s the different rationales that must be used when evaluating a material for its hemocompatibility. Next we e xplore some of the key point s of this important document. 2.2 Standard ISO 10993 4 The relevant standard for this research ISO 10993, defines blood/device interaction as any interaction between blood or blood components and a device resulting in effects on the blood or on any organ or tissue, or on the device [ 49 ] [ 50 ] When testing a material for its hemocompatibility, several considerations, such as dura tion of contact, temperature, sterility, flow conditions, sample collection/storage, enzymes, controls, surface to volume ratio and fluid dynamic conditions (flow rates), among others, have to be taken into account. Tests should be performed with minimal d elay, usually within 4 hours of blood withdrawal, since some properties of the blood change rapidly after collection [ 51 ] [ 50 ] Additionally, in the case of a specific geometry, the ratio of the test parameter to the exposed surface area must be evaluated (Concentration per unit volume/cm 2 ) [ 51 ] [ 50 ] [ 52 ] The use of animal models for screening purposes is encouraged, but the interpretation of the results must be done carefully, considering that results may be different in human models, given the dif ferences in blood reactivity between species, as well as the fact that differences between in vitro and in vivo model s will most probably exist. For example in vitro hemocompatibility does not guarantee in vivo hemocompatibility [ 49 ] [ 51 ] [ 53 ]


26 ISO 10993 4 provides a test selection system that is based on clinical concerns. Th e types of tests required by this standard depend on the blood contact category of the device or material (external communicating devices blood path indirect, external communicating devices circulating blood, and implant devices) [ 52 ] For each contact category tests are recommended from the following list of general test categories: thrombosis, coagulation, platelets/platelet function, hematology, and immunology, as shown in Table 2.1 for in vitro testing [ 52 ] Table 2 1 A sample of primary in vitro tests, Adapted from [ 51 ] Copyright remains with ISO Test Category Method Comments Thrombosis Adhered platelets, Platelet count, leukocytes, aggregates, erythrocytes, fibrin, etc. Light Microscopy or Scanning Electron Microscopy Coagulation Partial thromboplastin time Hematology Leukocyte count; hemolysis (plasma hemoglobin) Hemolysis is performed to measure the red blood cell membrane fragility in contact with materials and devices. Hematology Leukocyte count and differential; hemolysis (plasma hemoglobin) Hemolysis screening test is performed because of its measurement of red blood cell membrane fragility in contact with materials and devices. The method used should be one of the normative standard test methods for hemolysis. Immunology C3a, C5a, TCC, Bb, iC3b, C4d, SC5b 9 A panel including the last four tes ts encompasses the various complement activation pathways.


27 I n order to maximize the hemocompatibility information about a device, one or more tests from Table 2.1 should be performed. Unfortunately, in the absence of standardized and validated test methods, the requirement for testing across all categ ories is frequently ignored [ 52 ] Complement activation is the most important immunology test for circulating blood contact devices/materials; this test reveals the activation of the complement cascade [ 41 ] ASTM F 756, a standardized ASTM hemolysis test method, is used to determine the hemolytic potential of a device or material; an increase in plasma hemoglobin correlates with lysis of red blood cells, thereby indicating hemolytic activity of the material exposed to the cells. A device's effect on blood coagulation may be measured in vitro by determining the rate of clot formation or the partial thromboplastin time (PTT ) test of plasma exposed to the biomaterial or device during an incubation period. The reaction of white blood cells to materials can also be used as a n effective hematology test [ 52 ] However, the first step in determining the degree of hemocompatibility of a material is to evaluate thrombogenicity and com patibility with the cells of the blood vessels lining as low thrombogenicity and low compatibility with the vessel cells would not justify testing for hemolysis complement activation and immunologic responses or would require in vivo or ex vivo testing 2.3 Perspectives in Hemocompatibility A ssessment Previous research regarding the bio and hem ocompatibility of materials have shown interest in amorphous SiC, porous SiC, crystal line silicon, polysilicon Diamond like carbon (DLC), p olymers heparin coatings, titanium chromium and different oxides [ 8 ] [ 12 ] [ 35 ] [ 46 ] [ 54 ] [ 55 ] [ 26 ] [ 56 ]


28 As both Si and SiC have shown good bio compatibility for tissue proliferation they have drawn interest for hemocompatibility research, however, most of all the documented research regarding hemocompatibility of SiC is limited to amorphous SiC, 4H SiC 6H SiC and porous SiC ( 4H SiC and 6H SiC ) [ 57 ] Also, in many of the s e studies SiC is not compared directly to Si which is still the leading material for biomedical sensors To the best of our knowledge this is the first research that directly compares single crystal 3C SiC with Si (100), and explores if the low thrombogenic behavior of amorphous SiC extends to its single crystal (i.e., 3C SiC) polytype. Hemocompatibility of amorphous hydrogenated SiC (a SiC:H) has been tested mostly as a coating made by Plasma Enhanced Chemical Vapor Deposition (PECVD) [ 8 ] [ 55 ] these coatings have shown less protein adhesion and aggregation when exposed to whole blood than Si and silicon rich surfaces which tend to absorb more albumin [ 58 ] Chances of protein biofouling directly associated with the capacity of a material to adsorb proteins, and protein adsorption seems to decrease with increased concentrations of carbon. In [ 58 ] it was sh own than silicon and silicon nitride had higher platelet adhesion than s ilicon dioxide, paralene or SU 8 polymers unfortunately they did not include SiC in their study Polyethylene glycol has been grafted to silicon in the past t o decrease the protein binding and platelet adhesion of silicon surfaces [ 28 ] [ 59 ] [ 60 ] It has been shown that amorphous S iC coatings have low levels of fibrinogen and fibrin (proteins that promote clot formation) adsorption when contacting blood thus making SiC coatings, resistant to platelet adhesion [ 61 ] [ 21 ] Consistent results have been found when comparing amorphous SiC coatings to Ti. It is not known if the clot resistance of PECVD


29 deposited a SiC coatings will translate to single crystal 3C SiC film s grown on Si substrates In [ 54 ] a direct comparison of silicon wafers and a SiC coated (PECVD) silicon wafers was performed In this study both materials seemed to be outperformed by diamond like coated silicon wafers which presented less clot formation. Additionally, a microvascular approach has been introduced in many hemocompatibility studies, as this is the level where the cellular interactions responsible for hemocompatibility take place. T his micr ovascular approach is related to the assessment of microvascular endothelial cell proliferation on the surfaces of interest [ 21 ] [ 62 ] [ 63 ] [ 64 ] DLC and Si doped DLC films and coatings have been shown to have good compatibility with human microvascular endothelial cells (HMVEC) and d ecreased platelet adhesion. It was also shown than graphitiz ation of these films reduces HMVEC attachment and increases platelet adhesion [ 21 ] The evaluation of endothelial cell proli feration has been used as a standard to evaluate the hemocompatibility of vascular grafts, and it h as been used to evaluate Si and plasma modified polyethyl enerephthalate (PET) substrates, The study of the opposing but complementary nature of the behavior of platelets and endothelial cells for a particular biomaterial can therefore provide valuable insight about its hemocompatibility [ 21 ] From all the above, it can be concluded that the perspectives in the evaluation of the hemocompatibility of a material can be classified, at least for initial exploration, in to three types: 1 ) the evaluation of platelet adhesion and aggregatio n to the substrates of interest, 2 ) the evaluation o f the adsorption of the protein related to platelet adhesion to


30 the surf aces and 3) the study of microvascular endot helial cell proliferation on these surfaces. 2.3.1 The P latelet A dhesion and Activation Assessment P erspective The deadly complications associate d with the formation of a fibrin platelet thrombus (i.e. arterial or pulmonary embolization) when inserting a blood contacting device into the blood stream, or the risk of device malfunction make assessment of the behavior o f platelets in contact with a material a topic that is gaining increased interest. P latelet adhesion, lamellipodia and filopodia formation aggregation and spreading precede the formation of a thrombus and these markers have therefore been used as a icity [ 58 ] [ 65 ] [ 66 ] [ 67 ] [ 68 ] Figure 2 1 Scanning electron micrographs showing platelet morphology of Si (left) and SiO 2 (right). Images taken at 1200 x 0.15kV [ 58 ] Figure 2.1 illustrates, from the study in [ 58 ] the adhesion of platelets to Si and SiO 2 The images reve a l morphological changes of platelets (pseudopods formation) that indicate activation and aggregation.


31 F i gure 2.2 shows how in [ 58 ] platelet adhesi on was used to quantitatively compare various material used in Bio MEMS devices. The results confirmed higher levels of thrombogenicity for Si and the lowe st value of platelet adhesion for polyurethane [ 69 ] platelet circularity and spreading (area) have also been reported as more specific measurements of hemocompatibility. When a platelet activates it starts spreading on the surface and developing pseudopods, changing consequently its circular shape from the resting state. The less circular the platelets are that adhere to a surface, the more thrombogenic the material is [ 68 ] Figure 2 2 Number of adhered platelets per 15 mm 2 for Si, Si based materials, SU 8, parylene and polyurethane. # Statistically different from Polyurethane and stat istically different from Si [ 58 ] In contr ast with other blood formed elements platelets are not normally in stable contact with each other, but they develop such contacts once aggregation has started, limiting the escape of platelet activators from within the gaps between them T his helps to create and prevent the dissolution of the haemostatic plug [ 70 ] T herefore the degree of


32 aggregation also provides important information about the thrombogenic behavior of the material, even when activation is not present [ 71 ] 2.3.2 The Protein Adsorption Assessment P erspective W hen injury is present in the vessel wall, it initiates the expre ssion of tissue factor activity. Platelets rapidly accumulate on the endothelium, and fibrin can be observed at the platelet thrombus vessel wall interface, at the leading edge of the thrombus The same concept applies for thrombus formed on the surfaces of a material, with adhesion plasma proteins being adsorbed to the surface and fibrinogen being transformed to fibrin by the action of thrombin. Figure 2 3 F ibrinogen promoting platelet platelet adhesion by binding glycoprotein IIb IIIa (GPIIb IIIa). Adapted from [ 70 ] Fibrinogen and the Von Willebrand factor are thought to play a key role in platelet vessel wall interaction, particularly at low shear rates. Deficiencies in fibronectin delay ed the formation of thrombus in in vivo experiments performed on modifi ed rats [ 70 ]


33 The short distance between platelets leads to aggregation when by direct interaction a multivalent adhesive proteins (i.e. Fibrinogen) link activated glycoprotein IIb IIIa (GPIIb IIIa, the major platelet membrane integrin) on the surface of platelets adjacent to other platelets or promoting platelet adhesion to artificial surfaces (see Figure 2.3). However there are multiple plasma proteins and participants of this platelet surface (mater ial or vessel wall) interaction, including albumin and collagen that are directly implied in thrombus formation [ 70 ] [ 14 ] [ 61 ] [ 72 ] However in vitro application s look for the minimum amount of variables possible and usually limits the study of plasma protein interactions with surfaces to either fibr inogen or fibronectin [ 72 ] 2.3.3 The Endothelial C ell P roliferation P erspective Antithrombogenic polymeric coatings have been widely used to improve the hemocompatibility of materials [ 73 ] However a new trend is to coat the material with the natural Antithrombogenic surface of the human body: Vascular endothelial cells. As stated before the relationship between endothelial cells and platelets is opposite in nature but complementary. A material that has low platelet adhesion usually has very good endothelial cell proliferation, and these two requirements must be fulfilled for a material to be considered hemocompatible for real applications [ 74 ] Endothelial coatings have been tested in the past on modified DLC surfaces with the purpose of using it in vascular grafts and heart valve s but this requires the substrate in which the endothelial cells are been seeded to be compatible with these cells.


34 Figure 2 4 .SEM micrographs showing endothelial cells on three types of Cr DLC. Cr is 1,5 and 10% [ 62 ] Cr DLC compatibility with human vascular changes as a function of the percentage of chromium in the film has been studied in [ 62 ] Figure 2 .4 shows that DLC with 5% of Cr was the only for which endothelial cells presented good morphology. All the other percentages shown agglomerations, lack of spreading, and poor generation of lamellipodia and filopodia typical of HMVEC. Consistently, the surfaces with bad morphology pres ented lower proliferation. It was concluded that 5% Cr DLC is more hemocompatible than the other variations analyzed in this study. This is an example of how microvascular endothelial cell proliferation can be used to asses hemocompatibility. However, usin g only this information is incomplete, as g ood endothelial cell proliferation must complement low platelet adhesion for a hemocom patibility statement to be made.


35 Chapter. 3: Platelet Adhesion and A ctivation 3.1 The B lood and its E lements Blood is a bodily fluid that constitutes about 8% of the body 's weight, it circulates through the heart and vessels transporting oxygen and nutrients to the different tissues and organs of the human body and transporting waste products away from them. It is through a blood mechan ism called coagulation that the body heals itself and prevents excessive bleeding after injury Blood also executes many immunological (activation of the complement system that leads to inflammation and foreign body encapsulation) and regulatory (pH and t emperature) functions in the human body [ 75 ] [ 15 ] Blood is constituted of plasma (55%) and formed elements or organelles (45 % ). About 92% of the plasma is water and 7% proteins (albumin, globulin, fibrinogen, etc.) and the remaining are electrolytes gluc ose, amino acids, fatty acids, organic nutrients and waste [ 75 ] [ 15 ] The normal pH of arterial h uman blood is approximately 7.4. T he term blood serum refers to plasma without clotting factors (i.e., fibrinogen). Blood formed elements include erythrocytes (red blood cells), leucocytes (white bl ood cells) and thrombocytes (platelets). About 99% of the blood organelles are red cells [ 75 ] [ 15 ] Mature red cells in mammals do not have a nucleus. Red cells are produced in the bone marrow and usually lose their nuclei when they are released into the blood stream in order to provide more space for hemoglobin A fter being released red blood cells live


36 for about 100 120 days. Mammalian erythrocytes are flexible biconcave disks: flattened and depressed in the center, with a dumbbell shaped cross section (see Figure 3.2). A typical human erythrocyte disk has a diameter of 6 8 m and a thickness of 2 m, about one third of the vo lum e of a red cell is hemoglobin, responsible for the transport of about 98% of the oxygen in the human body [ 75 ] [ 15 ] Figure 3 1 Diagram of a red blood cell (erythrocyte) shape, in plan and c ross section [ 75 ] White cells or leucocytes are the cells of the immune system and there a r e at least five types of them. U nlike red cells, white cells retain their nuclei. White cells are divided into granulocytes (also known as polymorphonuclear leucocytes) and mononuclear cells. There are three types of granulocytes and two types of mononuclear cells, s ee Figure 3.2 Figure 3 2 Classification of the 5 types of white cells [ 75 ]


37 Neutrophils are attracted to sites of infection (chemotaxis), and destroy the micro organisms (mainly bacteria and fungi) responsible for the infection by ingesting them through a process known as phagocyt osis. Eosinophils target parasitic infections and also play a role in allergic reactions. Basophils have a role in immediate hypersensitive reactions and allergic and inflammatory responses. Mononuclear cells include lymphocytes and monocytes. Lymphocytes mediate immune responses and are clas sified in to three types: B, T and natural killers (NK). B lymphocytes secrete antibodies to work against antigens as part of the humoral immune response. T lymphocytes attack cells bearing foreign antigen and antibody coated cells playing a key role in the cell mediated immunity. NK c ells reject tumors and attack cells infected by viruses; they play a role in cell mediated immunity and are a major component of the innate immune system. Monoc ytes migrate to tissues where they differentiate into tissue resident macrophages or dentritic c ells. Macrophages phagocytose cellular debris and pathogenic microorganisms and stimulate lymphocytes and other immune cells that respond to the pathogen. Dentritic cells present the antigens to the T lympho cytes in order to activate them. Another formed e lement in the blood are platelets Platelets are formed by the fragmentation of the protruding cytoplasm of megakaryocytes (large bone marrow cells) in the sinusoids of the bone marrow. Therefore they are not cells but fragments of cell cytoplasm of discoid shape The average life span of a platelet is 7 days and they range from 2 m to 3 m in diameter by 0.5 m in thickness [ 76 ] Their small size and discoid


38 shape allows the blood flow to push them to the vessel edge, near the surface of the endothelium were they are ready to detect and respond to vascular damage. Platelets play a key role in hemostasis leading to the formation of blood clots when in contact with activating factors C lot format ion requires platelets to undergo rapid morphological change s from a resting state to an activated state, as well as platelet aggregation, as shown in Figure 3.3. Figure 3 3 S canning electron microscope images of resting (left), partially (center) and fully activated (right) human platelets. From [ 77 ] The shape of the platelets is maintained by a sturdy and flexible interna l cytoskeleton composed of actin and tubulin polymers and their associated proteins [ 5 ] Remodeling of the cytoskeleton and assembly of new actin filaments allows for platelet shape changes. The surface of the platelets is featureless except for semi selective conduits formed from membrane invaginations that allow small molecules to enter the platelet membrane. Platelet cytoplasm is composed of normal cellular organelles (lysosomes, mitochondria, etc.) and additionally c ontains specific platelet granu les granules, which store


39 matrix adhesive proteins and have glycoprotein receptors in their limiting membranes in order to promote adhesion between platelets and the matrix. When blood is collected in tubes containing anticoagulant solution (otherwise it would create a single solid clot ) and it is allowed to settle it separates in to three layers see Figure 3.3 : The bottom layer is composed of red cells, the top layer is clear and pale yellow and is the blood plasma In between is a narrow layer called the buffy coat composed mainly of white cells and platelets [ 75 ] [ 15 ] I f it is necessary to isolate platelets from whole blood, two major methods are available: the platelet rich plasma (PRP) and the buffy coat (BC) methods, illustrated in the diagrams of Figure 3.4. Figure 3 4 Diagram of a tube of anticoag ulated blood that has settled illustrating the separation of blood into three layers. Adapted from [ 75 ]


40 Figure 3 5 Preparation of platelet concentrates using the PRP and BC techniques [ 15 ] Both techniques are based on the concept of differential sedimentation, which is manipulated via centrifugation. The sedimentation rate is influenced heavily by the physical properties of the cells and the viscosity of the medium, in this case, plasma [ 15 ] In the PRP technique, anticoagulated whole blood first undergoes low G force ( "soft" ) centrifu g a tion which results in a PRP supernatant that contains mostly suspended platelets( plus 30 50% of whi te cells and some red cells). The red cells form a pellet in the bottom of the test tube with the remaining white cells and some platelets. The PRP can be centrifuged at high G, to separate platelet poor plasma from the platelet concentrate pellet that can be resuspended to the desired density [ 15 ] High speed (hard) centrifugation separates blood differently from l ow speed centrifugation. A t high speed, the white cells sediment with red cells initial ly, leavi ng in the top a PRP supernatant. T hen red cells ar e packed closely and fall to bottom rapidly, which forces white cells upward at the plasma interface, which is a process that is followed by platelet sedimentation at the white cell layer. The form er describes the first


41 step of the BC method. Subsequently the buffy layer containing white cells and platelets is centrifuged at low g, resulting in dense platelet rich plasma [ 75 ] [ 15 ] Appropriate selection of the centrifugation speed and centrifugation time may result in a PRP supernatant, an intermediate white cell layer and a red cell pellet. 3.2 Platelet Adhesion and A ctivation Assessment P rotocol Platelet reactivity, measured by the degree to which platelets in suspension adhere to a surface in vitro has been proposed as a measure of thrombogenicity [ 58 ] This includes the a mount of platelet adhesion as well as changes in platelet morphology including increased area and development of pseudopods. Platelet circularity, C, is defined as where P is the perimeter of the platelet and A is its area. When the platelet possesses per fect circularity C is equal to unity [ 78 ] [ 58 ] It has been observed that an increase in platelet circularity on a material correlates with increased hemocompatibility of the material surface [ 58 ] It has also been noted that an increase in platelet area represents a decrease in hemocompatibility of the surface (this is cons istent with the above as the platelet tends to form a tight circular shape when not adhering to the surface thus minimizing its area). Therefore platelet spreading (i.e., increase in platelet area) and circularity have b een reported as specific markers for surface hemocompatibility. In order to evaluate platelet reactivity on 3C SiC, 4H SiC, 6H SiC and Si samples the following protocol was followed. Blood samples were obtained from young, healthy, female farm pigs weighting approximately 45 kg, free of hepatitis or other blood borne disease. The blood samples were collected by routine phlebotomy using 4.5 ml BD


42 Vacutainer blood collection tubes (blue stopper) containing 0.105M (3.2%) buffered sodium citrate (12.35mg of sodium citrate / 2.21mg of citric acid) for anticoagulation. The anticoagulant effect of sodium citrate is reversible, allowing the platelets to activate when in contact with a thrombogenic (i.e., non hemocompatible) mat erial. Platelet rich plasma ( PRP ) was obtained from the blood samples via centrifug ing the tubes for 20 minutes at 200 g and at room temperature The platelet concentration was calculated using bright line counting chamber (hemocytometer) with reduced thick ness latelets and leukocyte counts. T he average platelet concentration was 4 .38 x 10 5 platelets/ml The PRP samples were then incubated for 30 minutes at 37C to allow the platelet s to recover their shape which is necessary for platelet interactions and activation. Single crystal 3C SiC samples grown on Si(100) and single crystal Si(100) wafer samples ( die size 10mm x 8mm) were cleaned for 10 minutes via piranha solution (organics removal) at 2:1 concentrations of H 2 SO 4 : H 2 O 2 then rinsed three times with de ionized water (DI water) with followed by 5 minutes of H 2 O:HCl in a 1:1 ratio, intended to eliminate any sulfur traces resulting fr om the piranha cleaning. Native surface oxide was removed by dipping the samples in a hydrofluoric acid solution (50:1 H 2 O:HF) for 1 minut e The samples were then thoroughly rinsed in DI water and stored under ethanol. Just prior PRP exposure, the samples were rinsed three times with DI water and manually blow dried with nitrogen 4H SiC and 6H SiC samples were obtained from wafers purchased from Cree Inc. 4H SiC, 6H SiC and Si samples underwent the same cleaning procedure already described for 3C Sic samp les.


43 The clean and dry semiconductor material samples were them placed in BD Falcon 24 multiwell culture inserts and covered with 2 ml of PRP. The multiwell plates were then placed in the incuba tor for 15 minutes at 37C and 5% CO 2 and 95% relative humidit y The wafer die were then removed and rinsed t wice with PBS ( Phosphate buffered saline) to remove non adherent platelets, and then the pla telets were fixed using 4 % paraformaldehyde in PBS The samples were rinsed again two times and then dipped for 30 minutes i n a 0.1% Triton X solution in order to permeabilize the membrane of the adherent platelets. After rinsing, the fixed platelets were pre incubat ed in PBS containing 1% BSA ( bovine serum albumin) in order to reduce nonspecifi c backgro und staining. Next, the F Actin filaments of the platelets' cytoskeleton was stained using Rodhamine Phalloidin a phalotoxin conjugate d to the orange fluorescent dye, tetramethylrhodamine (TRITC ) also diluted in 1% BSA solution Cover s lips were mounted over the samples using DAPI ( 4',6 diamidino 2 phenylindole) a nucleic acid fluorescent stain, in order to differentiate platelets from any unintentionally present cell s derived from the PRP separation process. Subsequently, the samples were imaged using a Leica DM IL inverted microscope at different levels of magnification, and the images were manually processed to enhance borders and backgrounds and then automatically p rocessed (thresholding, noise reduction, etc.) and analyzed using ImageJ, a public domain image processing software developed at the NIH (National Institutes of Health) [ 79 ] Previous research regarding plate let adhesion, and activation were usually performed in the absence of aggregation, either chemically manipulating the PRP (for example


44 adding EDTA) or by exposing the surfaces to PRP only for 5 minutes [ 58 ] [ 80 ] A dditionally the PRP is diluted to guarantee a low number of platelets adhered to the surface and this way images of single activated platelets are easier to obtain [ 81 ] Although this approach allows the acquisition of single activated platelet images it may be detrimental to the experiments as it can yield falsified results making it hard to know if low adhesion and activation is just a consequence of low platelet surface interaction, it also prevents one from getting a better picture of platelet clumping in the presence and absence of activation [ 81 ] In this experiment the samples were exposed for longer times (15 minutes) at high platelet concentrations and without PRP manipulation to avoid aggregation, in order to get a better insight in to the thrombogenic behavior of the studied surfaces S ubsequently these conditions resulted in a higher level of difficulty when analyzing the images, as the presence of clumps and the fact s that platelets do not have a constant size and that their shape is discoid required more complex image processing (manual and aut omatic) in order to facilitate particle differentiation when possible ( see Figure 3.6 ) Original Image Manual edition Thresholding and eroding Automatic particle detection and analysis Figure 3 6 Illustration of the image analysis process for platelets seeded on a Si sample. The original image cannot be adequately processed by the software. The image is first manually enhanced and then automatically prepared for particle detection and analysis wi th ImageJ [ 79 ]


45 Higher platelet adhesion was observed on the Si surface than on SiC, and among the SiC polytypes it can be seen that more adhesion occurred for the hexagonal types than the cubic (i.e., 3C SiC) type. More importantly, when looking at high magnification it was found that the platelets that adhered to the 4H SiC and 6H SiC surfaces had more activation (presence of lamellipodia and filopodia) and presence of clumps than for 3C SiC and Si. 3C SiC surfaces showed less aggregation and activation with much more c ircular shaped platelets that the other surfaces thus indicating an improved hemocompatibility. The for mer is illustrated in Figure 3.7 Si(100) 3C SiC/Si(100) 4H SiC 6H SiC 40x 100 x Figure 3 7 Fluorescent micrographs c omparin g adhered platelets on 3 different SiC polytypes and Si. 3C SiC clearly shows lower adhesion and high circularity while Si presented the highest adhesion. 4H SiC and 6H SiC evidenced the pres ence of clumps and high degree of activation 4H SiC presented, as shown in Figure 3.8 ., large area activated particles and non activated/activated platelet clumps and the highest adhesion of all the SiC polytypes included in this study, exhibiting therefore a thrombogenic behavior that can lead to device failure.


46 100x Figure 3 8 Fluorescent micrographs revealing activated platelet clumps (left), non clumped activated platelets (center) and a large non activated platelet clump on the surface of 4H SiC samples. Magnification 100x The results of these experiments were evaluated for statistical significance using analysis of variance (ANOVA) to detect significant differences among all pairs of means of the markers circularity, particle adhesion counts, and particle size (this will indicate if at least one pair of mean values is statistically different) H owever this technique does not single out any particular pair of means a s being statistically different. I n order to address this problem for the cases in which the F test in dicated that at least one of the means was different a multiple comparison using paired T tests with Bonferroni adjustment was used. The tests were evaluated for a significan D ue to the presence of platelet aggregation the resulti ng statistics cannot be interpreted in a deterministic way in order not to misinterpret particle analysis For example a very circular clump where platelets cannot be differentiated is going to yield a big size (false platelet size) and in this case c ircu larity is not significant compared to size, that indicates the presence of clumps S imilarly, a low number of counts, but high size indicates the presence of a few big clumps, where individual platelets cannot be differentiated automatically by the softwa re. A low thrombogenic material should present low counts of adhered platelets with high circularity and small average size.


47 Bar diagrams and ANOVA analysis tables for circularity, adhesion and platel et area (size) are shown in F igures 3.9, 3.10 and 3.11 respectively. Platelet circularity for all the samples was above 0.85 which is considered a high value and indicative of low activation [ 78 ] T he highest average value was for 3C SiC with a confidence in terval of [0.939 + 0.049] also with the smallest standard deviation. H owever when testing for statistical differences a high p value of p =0.0746 showed no significant differences among the platelet circularity for the different material s T his also suggests that the samples have similar variances and therefore the presence of out liers in the circularity measurements for 4H SiC. This indicates that the results of circularity, particle counts and size should be analyzed together to understand the throm bo genic behavior of this material. 4H SiC presented high circularity but small particle count (comparable to 3C SiC and 6H SiC, as confirmed by paired adjusted test), particles that also presented the highest average size with the highest variance. T his ca n be interpreted as the particles not being individual platelets, but large size circular clumps that the software is not able to differentiate as multiple particles regardless of any image enhancement done previously leading to platelet miscounts Looki ng at the circularity alone, it would lead us to think of 4H SiC as a low thrombo genic material however this cannot be further from the truth, and a composite analysis of the statistical data confirms the optical results illustrated in Figures 3.7 and 3.8 Additionally, it can be concluded that mathematically small differences in circ ularity a re good indicators of the initial stages of activation, even though they may not be big enough to translate into drastic statistical indicators of variability among the samples F or example the confidence intervals of circularity for 3C SiC and Si were [0.939 + 0.049]


48 and [0.885 + 0.053] which is a small difference with no statistical significance, but optical comparis on of Si and 3C SiC (F igure 3.12 ) shows that platel ets adhered to Si are more activated than the ones adhered to 3C Si C, however they are in an early stage of activation which must be considered when drawing conclusions from this data 6H SiC or 4H SiC had comparable average circularity but also higher var iability, which was associated either with the presence of circular clumps or with the prese nce of deformed clumps or indivi dual highly activated platelets Statistical paired analysis of the particle count s indicated no statistical differences between 3C SiC and 6H SiC ; however platelets that adhered to 3C SiC showed higher circularity and smaller area while 6H SiC evidenced the presence of activated clump s Si surfaces evidenced the highest platelet adhesion of all the surfaces with high presence of clumps (large size particles) thus showing consistently a lower level of hemocompatibility.


49 Figure 3 9 Statistical analysis of particle circularity on 3 different SiC polytypes and Si. Top: Histogram Bottom: Anova table. Figure 3 10 Statistical analysis of particle counts on 3 different SiC polytypes and Si. Top: Histogram Bottom: Anova table.


50 Figure 3 11 Statistical analysis of particle size on 3 different SiC polytypes and Si. Top: Histogram Bottom: Anova table. Si(100) 3C SiC/Si(100) 100x Figure 3 12 Fluorescent micrographs compari n g Si (left) and 3C SiC (right) samples revealing higher degree of thrombogenicity in Si than in 3C SiC. Si samples presented higher adhesion, activation and aggregation than 3C SiC samples.


51 3.2.1 The Impact of Surface R oughness Surface roughness is considered a key factor in determining a low thrombogenic behavior of biomaterials. Although the mechanisms behind hemocompatibility have not been completely understood yet, it is l argely accepted that blood cell lysing due to high sur face roughness can induce a thrombogenic reaction but more importantly that surface roughness has an impact o n surface wettability and protein adsorption [ 82 ] [ 83 ] [ 34 ] In order to assess t he effect of surface roughness o n platelet reactivity to 3C SiC several samples of high and low roughness were used, with different degrees of cry stal linity and crystal orientation: single crystal 3C SiC samples grown on Si(100) showing low roughness Polycrystalline 3C SiC grown on Si(100) showing high roughness, and single crystal 3C SiC grown on Si(111) evidencing both high roughness and low roughness in different parts of the wafer, see F igures 3.13 3.17 Table 3.1 summarizes the range of values used for roughness assessment of the different samples. The silicon samples were u sed as controls for comparison and its roughness has average value s smaller than single c rystal 3C SiC and a small standard deviation, indicating less variability of the surface morphology. Table 3 1 Summary of samples roughness expressed as [ average roughness ( ) + s tandard deviation of the mean( )] in nanometers (nm) Sample [ + ( nm ) Single crystal 3C SiC/(100)Si 0.248167 + 0.112642 Polycrystalline 3C SiC/(100)Si 62.31167 + 17.38315 Single crystal 3C SiC/(111)Si ( low Roughness ) 4.215833 + 1.525045 Single crystal 3C SiC/(111) Si ( high roughness ) 34.0165 + 16.6633 (100) Si 0.074 + 0.087681


52 Figure 3 13 AFM micrograph taken in tapping mode. Example of sample with low roughness: single crystal 3C SiC/(100)Si (from wafer USF1 08 013 die #9) Figure 3 14 AFM micrograph taken in tapping mode. Example of sample with high roughness: polycrystalline 3C SiC/(100)Si (from wafer USF2 06 074 die #4)


53 Figure 3 15 AFM micrograph taken in tapping mode. Example of sample with low roughness: single crystal 3C SiC/(111)Si (from wafer USF U die #9) Figure 3 16 AFM micrograph taken in tapping mode. Example of sample with high roughness: single crystal 3C SiC/(111)Si (from wafer USF U die #4)


54 Figure 3 17 AFM micrograph taken in tapping mode. Silicon (100) sample exhibiting low roughness. The protocol used to evaluate the effect of surface roughness was the same as described in S ection 3.3. After fluorescence microscopy of the stained platelets at different levels of magnification, it was observed that there is a high density of clumps for the high r oughness surfaces that is independent of crystal orientation and surface morpholog y Figure 3.19 Low roughness surfaces on the other hand evidenced in teresting results, as shown in F igures 3.1 8 and 3.20. Figure 3.18 compares two low roughness surfaces with different orientation s the optical results indicate that for comparable rough ness, crystal orientation may have an impact on platelet reactivity Figure 3.20 shows high platelet adhesion, activation and clump formation on a low roughnes s Si surface; comparison of these images wit h the left images in Figure 3.18 indicate higher reac tivity of Si versus 3C SiC, this is, for identical conditions (both single crystal, same orientation and low roughness) which can be explained by the fact that silicon surfaces are considerably mor e


55 hydrophobic than the SiC surfaces [ 9 ] ; it is a well known fact that adhesion proteins adhere preferentially to hydrophobic surfaces [ 84 ] Single crystal 3C SiC/Si (100) Single crystal 3C SiC/Si (111) 40x 100 x Figure 3 18 Fluorescent micrograph c ompari son of two low roughness surfaces. Left: single crystal 3C SiC/Si(100). Right: single crystal 3C SiC/Si(111)


56 P oly crystalline 3C SiC/Si( 1 00 ) Single crystal 3C SiC/Si( 111 ) 40x 100 x Figure 3 19 Fluorescent micrograph c omparison of two high roughness polycrystalline 3C SiC surfaces. Left: Grown on a Si(100) substrate. Right: Grown on a Si(100) substrate 40x 100x Figure 3 20 Fluorescent micrograph of p latelet adhesion on a low roughness silicon (100) surface


57 Unlike the previous experiment, no particle analysis was derived from the images, as the high density of adhesion and aggregation for all the surfaces, except for the single crystal 3C SiC/Si(100) would only result in falsified statistics. 3.3 Discussion Pla t elet adhesion a nd morphology were assessed after the exposure of PRP to different SiC polytypes different crystal orientations and surface roughness es and the results were contrasted to those of Si (100) wafer die The information obtained from these experiments was evaluated qualitatively, by optical (fluorescent microscopy) inspection, and quantitatively, by obtaining specific information from the images and then performing statistical analysis. In each case sing le crystal 3C SiC grown on Si(100) substrates had the best performance, with low particle adhesion counts and small area adhered platelets that also showed high platelet circularity. The former characteristics are highly desirable for cardiovascular applic ations as they display low thrombogenicity and reduce the chances of catastrophic failure of a device due to thrombi embolization.


58 Chapter. 4: Protein Adsorption to S urfaces 4.1 The Aggregation and A ctivation of P latelets Platelet aggregation is a key event in the process of thrombus formation, which results from a series of coordinated cell adhesion events. Circulating platelets are initially tethered to subendothelial matrices exposed at the vascular injury site, followed by stable platelet adhesion to the subendothelium, and a dherence of additional platelets to the surface of platelets already adhered leading to platelet aggregation. This sequence is mediated by specific receptor ligand interactions [ 85 ] [ 5 ] IIb 3 (the same as glycoprotein [GP] IIb IIIa complex) is the most abundant platelet surface membrane glycoprotein and its expression is specific to platelets and megakaryocytes. This glycoprotein is essential for platelet aggregation as this event is mediated b y the binding of fibrinogen or v on Willebrand F actor ( v IIb 3 (see Figure 4.1). IIb 3 is a heterodimeric receptor th at mediates platelet adhesion and signaling by modulating their affinity and avidity for adhesive ligands; this process is known as integrin activation or inside out signaling [ 5 ] Platelet aggregation is a form of cell to cell adhesion that is mediated by the binding of adhesive ligands to GPIIb IIIa. Even though platelet activation and aggregation is induced by a variety of stimuli, the aggregation response specifically requires t he presence of GPIIb IIIa. Following that logic, GPIIb IIIa receptor antagonists have been used successfully in clinical trials as antithrombogenic agents [ 72 ] [ 86 ]


59 GPIIb IIIa plays a fundamental role in the process of thrombus formation, either physiological (as a response to injury) or pathological (thrombosis). Figure 4 1 Platelet aggregation mediate d by the action of fibrinogen and von Willebrand Factor ( vWF ) [ 87 ] GPIIb IIIa is expressed on the plasma membrane of the resting platelets, the granules and the open canalicular system as Ca 2+ dependent heterodimers. Additionally, it can bind to several macromolecular adhesive gl y coproteins (e.g., fibrinogen, vWF, fibronectin, and vitronectin, prothrombin and neural cell adhesion molecule L1). Arginine Glycine Asparti c acid, known as RGD, is a sequence present in the ligands that bind to GPIIb IIIa, that serve as a recognition motif promoting cell to cell and cell to matrix adhesion [ 88 ] In resting platelets GPIIb IIIa has a low activity for ligand binding, but when platelets are exposed to a soluble antagonist (e.g. thrombin) or subendothelial matrices, their ability to bind ligands increases rapidly. Stimuli engage specific membrane receptors that induce cy toplasmic intracellular signals that activate GPIIb IIIa. The ligand necessary for platelet aggregation in vitro is fibrinogen [ 72 ] [ 5 ]


60 4.2 The QCM T echnology The quartz crystal microbalance ( QCM ) is basically a mass sensing device with the ability to measure very small mass changes on a quartz crystal resonator in real time. The sensitivity of the QCM is approximately 100 times higher than an electronic fine balance with a sensitivity of 0.1 mg [ 89 ] This means that QCMs are capable of measuring mass changes as small as a fraction of a monolayer or single layer of atoms [ 90 ] Figure 4 2 Diagram of a quartz crystal sensor [ 91 ] The high sensitivity and the real time monitoring of mass changes on the sensor crystal make QCM a very attractive technique for a large range of applications including its use under vacuum and more recently its use with fluids or with viscoelastic deposits. The QCM principle of operation is based on a piezoele ctric quartz crystal sandwiched between a pair of electrodes (see Figure 4.2) [ 90 ] [ 89 ] When the electrodes are connected to an oscillator and an AC voltage is applied over the electrodes the quart z


61 crystal starts to oscillate at its resonance frequency due to the piezoelectric effect. This oscillation is generally ver y stable due to the high Q factor If a thin, non dissipative layer is evenly depos ited on the active electrode the resonant frequency will decrease proportionally to the mass of the adsorbed layer according to the Sauerbrey relationship [ 90 ] f is linearly related to the adsorbed mass = (Eq. 1.) where C is a constant based on the physical properties of the quartz crystal ( = 17 7 2 at f =5MHz) an d n is the overtone number ( n =1,3,5,...). If the damping of the deposited layers can be con sidered to be negligible, then the QCM is used in order to study the dissipative processes on the quartz crystal. The Sauerbrey relatio n ship is not applicable for hydrated, highly viscous surface adsorbed protein layers that result in high dissipation shif ts [ 90 ] In those cases, i n addition to measuring the frequency, the dissipation is often measured to help with the analysis and the techinique is called QCM D, where the D stands for dissipation The dis sipation is a parameter quantifying the damping in the system, and is related to the sample's viscoelastic properties. Theorethical modeling of the respo nse using dissipation requires th e use of a Voigt model, that characte rizes the adsorbed layer using th e following parameters: effective density [ 90 ] [ 72 ] This method has been used mainly to study the interactions of proteins (including blood serum proteins) with film coatings, especially polymeric coatings [ 88 ] [ 92 ] [ 93 ]


62 4.3 Protein Adsorption Assessment U sing QCM D Platelet aggregation is the primary mechanism of haemostatic control. It is widely accepted that the hemocompatibility of a material is highly dependent of platelet activation and aggregation onto its surface, which at the same time is dependent upon the binding of fibrinogen to the surface of the material and of ac tivated platelets which is mainly mediated by platelet receptor GPIIb IIIa [ 85 ] Adsorption of plasma proteins to ma terials influences the plasmatic (c oagulation and complement) and cellular activation pathways [ 72 ] [ 94 ] These adhesive proteins, particularly fibrinogen adsorb to surf aces within seconds of contact. It has been reported in [ 95 ] that platelet adhesion to surfaces, in vitro and preadsorbed with plasma is mediated e xclusively by fibrinogen and does not involve other adhesion plasma proteins. The refore a better insight into fibrin ogen adsorption to surfaces and platelet membrane receptor binding efficiency will provide a better understanding of the hemocompatibility of a material. QCM D can monitor the formation of protein layers adsorbed onto the surface of a quartz crystal in rea l time, by monitoring the resonant frequency shifts In this study we monitored the binding kinetics of purified GP IIb IIIa to adsorbed fibrinogen using the methods and protocols described in [ 72 ] and [ 96 ] In order to assess the protein adsorption to 3C SiC and compare it to Si, custom made sensors were ordered from Q Sense [ 91 ] We obtained four SiC coated sensors, four gold (i.e., standard QCM D surfaces) four silicon coated and four nanocrystalline diamond (NCD) coated sensors. The gold sensors were used as sample controls, as it is known that gold is used for blood contacting applications, and the NCD sensors to start


63 exploring the hemocompatibility of th is material due to the interest generated after Dr. Frewin's research [ 11 ] The main purpose was to compare fibrinogen and GPIIb IIIa adsorption to 3C SiC vs. Si. Protein free TRIS buffer ( Tris (hydroxymethyl)aminomethane ) was prepared using Trizma HCL (cat. 1185 53 1), Trizma Base (cat. 77 86 1) from Sigma Aldrich and d e ionized water (with a to obtain a 50mM solution with a 7.4 pH at 37C. D ivalent cations were added to the TRIS buffer as they are required for GPIIb IIIa binding to ligands [ 72 ] [ 5 ] [ 97 ] as follow s : 100 mM NaCl (Sigma Aldrich, cat. S5886 500G), 1 mM MgCl (Sigma Aldrich, cat. M8266 100G), 1 mM CaCl 2 (Fisher, cat. S71924) and 1 mM MnCl 2 (Fisher, cat. M87 500). Human fibrinogen (cat. 341575) and human platelet receptor GPIIb IIIa (cat. 528240) were obtained from Calbiochem and prepared to the following final working concentrations: fibrinogen 50 g/mL and GPIIb IIIa 10 g/mL. Bovine serum albumin (BSA) was obtained from Fisher Scientific (Cat. BP671 10) and prepared at a working concentration of 10% BSA (w/v). BSA was used to block nonspecific GPIIb IIIa binding to the surfaces. Hydr ogen peroxide (cat. 216763 100 ) and ammonium hydroxide (cat. 320145 500M L) were obtained from Sigma Aldrich and used to clean the quartz crystal sensors (Au, Si, SiC and NCD) following the protocol A I from Q Sense. 2% (w/v) of sodium dodecyl sulfate (SDS) from Sigma Aldrich (cat. L4390 100G) was prepar ed in DI water in ord er to rinse extensively the Q sense instrument chambers before and after every experiment.


64 Fibrinogen adsorption and the corresponding binding of the platelet receptor (GPIIb IIIa) to quartz crystals with a fundamental resonant frequency of 5 MHz coated with Si, SiC and NCD was measured Frequency shifts versus time curves, induced by sequential adding of protein solutions (human fibrinogen, BSA, human GPIIb IIIa), were recorded. Each protein solution was incubated until the binding saturation was reached followed by a TRIS buffer rinse in order to remove non bound proteins All experiments were repeated six times using the Q Sense E4 monitoring system at 37C, in TRIS buffer under non flow conditions and allowing the liquid in the QCM D chambers to exch ange in every step via gravitational flow. 1.5 mL of solution per chamber (TRIS, fibrinogen, BSA, GPIIb IIIa) was loaded in every step and was sufficient to fill the chambers and to allow for fluid exchange. All the solutions were degassed before exposure to the chambers and crystal surfaces. Frequency shifts induced by adsorbed proteins were monitored in real time at three, five, and seven However the data provid ed by the first overtone had better stability higher sensitivity and less glitch like errors thus it was the one used for subsequent statistical analysis. In order to obtain the baseline for the observed frequency shift s 1.5 ml of degassed TRIS was allowed to sit in each chamber. Once a constant b aseline was obtained (and the values compared to the acceptable values provided by Q Sense) the sensors were exposed to the fibrinogen solution until saturation was reached, then it was followed by TRIS rinsing and BSA incubation for 30 minutes. Next, the BSA was rinsed with a TRIS buffer and the chambers filled with GPIIb IIIa. The solution was allowed to sit until saturation was reached. The chambers were rinsed thoroughly again using TRIS (see


65 Figure 4. 3). The measurements were stopped and the sensors a nd chamber cleaned using SDS Figure 4 3 Sequence of steps of a typical hemocompatibility assesment using a Q sense device. TRIS buffer (50 mM TRIS, pH7.4 at 37C containing 100 mM NaCl, 1 mM IIIa 10 Adapted from [ 96 ] After every experiment the senso rs were cleaned (Q Sense's cleaning protocol A I) using an UV/Ozone chamber for 10 minutes, followed by dipping the sensor in a H 2 O 2 (30%), NH 4 OH (25%) and DI water solution in a 1:1:5 ratio for 5 min at 75 C. The sensors were rinsed with DI water, dried using nitrogen gas and stored in a desiccator, to avoid surface changes due to contact with contaminated air. 4.4 Results The custom made coated quartz sensors were create d by sputtering the material of the desire d coating onto the surface of existing gold sensors. Unfortunately the NCD films


66 peeled off from the gold surface after the cleaning and no measurement s were obtained for those films Additionally, after XRD, it was found that the SiC sensors did not corr espond to 3C SiC (as it was requested) but to 4H SiC. This unfortunate event did not allow the direct comparison of 3C SiC versus Si that was desired for this experiment. Figure 4.4 shows a histogram of the frequency shifts due to the loading of fibrinogen and GPIIb IIIa for gold, silicon and 4H SiC. From the statistical results (expressed as the mean + standard deviation) it is evident that there is much more fibrinogen adsorption on Si and 4H SiC than on gold and the opposite is observed for GPIIb IIIa. In order to confirm these results ANOVA analys is (se e Figure 4.5 ) and paired T tes t with Bonferroni adjustment were performed at a confidence level of 95 % and it was confirmed that th e behavior of 4H SiC was not statistically different from silicon Gold instead displayed significant differences from 4H SiC and Si. Figure 4 4 Histogram of the fribrinogen and GPIIb IIIa adsorption to gold, silicon and 4H SiC substrates. Measurements made using a Q Sense E 4 QCM D tool in the USF FCoE BITT laboratory [ 98 ]


67 ANOVA table for f ibrinogen related frequency shifts ANOVA Table for GPIIb IIIa related frequency shifts Figure 4 5 ANOVA tables of the statistical data for Fibrinogen (top) and GPIIb IIIa (bottom). Notice the low p value that indicates that statistical differences exist in at least one pair of means. Interestingly, 4H SiC and Si presented high levels of Fibrinogen ad sorption to their surfaces. H owever, the percentage of GPII b IIIa that bound to the adsorbed fibrinogen was less than the one that bound to gold. The purpo se of these experiments was to determine which material had higher degree of protein adsorption. The substrates used were hard and non viscous and the observed dissipation values were negligible (the maximum value observed was 3.4 x 10 6 for GPIIb IIIa on gold,) and therefore Voigt models were not derived. Instead the mass adsorption (expressed in ng/cm 2 ) was derived using the Sauerbrey relationship (See F igure 4.6).


68 Figure 4 6 Histogram of the fribrinogen and GPIIb IIIa mass (ng/cm 2 ) adsorption to gold, silicon and 4H SiC substrates Measurements made using a Q Sense E 4 QCM D tool in the USF FCoE BITT laboratory [ 98 ] 4.5 Discuss ion From the previous platelet assessment experiment, we know that 4H SiC performs similarly to Si. These results were confirmed by the protein adsorption assessment using QCM technology. Unfortunately this experiment did not provide the desired informat ion regarding hemocompatibility of 3C SiC. Additionally, as more of the fibrinogen that adhered to gold mediated GPIIb IIIa bin ding, this result also implies a high er thrombogenic behavior of the gold surfaces used in these experiments Further experiments involving the interaction of platelets with these protein layers in order to determine if these values are directly correlated with higher platelet aggregation to these surfaces in vitro are required. The small values for dissipati on observed in these exper iments contradict the results reported in [ 99 ] w h ere the dissipation was not negligible, and because the substrates for


69 those experiments were polymeric films, th at most likely were being hydrated as the experiments were performed.


70 Chapter. 5: Microvascular Endothelial Cell Proliferation on Semiconductor S ubstrates 5.1 Microvascular Endothelial Cells and the Vessel Internal L ume n Blood supply to the different tissues in the body depends on the blood vessels. The internal lining of the blood vessels is composed of endothelial cells the natural non thrombogenic surface of the human body. If the continuity of the endothelial layer is disrupted, this event initiates the a ctivation of the clotting cascade. Endothelial cells form the interface between the cells in the lumen and th e vessel wall creating a flat pavement like pattern that seals the vessel [ 100 ] Endothelial cells are very flat, have a central nucleus, are about 1 2 m thick and some 10 20 m in diameter [ 100 ] as shown in Figure 5.1. Figure 5 1 Fluorescent micrographs illustrating the pavement like pattern of the endothelial cells. Endothelial cells were stained using CMFDA cell tracker dye.


71 Endothelial cells are considered an adaptable life support system, extending cell migration to almost every region of the body and adjusting their number and configuration to adjust to local requirements [ 100 ] This adjustable capability of the endothelium the thin single sheet of endothelial cells that lines the vessel wall, is responsible for blood vessel remodeling, growth, angiogenesis (formation of new blood vessels) and repair [ 100 ] Endothelial cells ac t like selective filters which regulate the passage of gases, fluid s and various molecules across their cell membranes th r ough the action of their cell surface molecule receptors and interaction sites that usually attract leucocytes to the endothelium in t he presence of inflammatory states Many leucocytes pass through endothelial cells, especially in capillaries as part of their normal life cycle, so that they can monitor foreign agents (antigens) in the tissues. Macrophages travel in the blood and pass through endothelial cells to gain access to various tissues of the body [ 100 ] [ 101 ] Endothelial cells also allow large amounts of blood plasma fluids t o pass through their intercellular junctions and create edemas as response histamine released after tissue damage. Another vital function of the endothelial cells is the synthesi s of the Vo n Willebrand's f actor e ssential for h emostasis [ 101 ] Large blood vessels (i.e., arteries and veins) have a thick wall of connective tissue and many layers of smooth muscle. The lumen of the large vessels is lined by the endothelium, which is separa ted from the outer layer by a basal lamina a layer of extracellular matrix found on the basal surface of epithelial cells and secreted by them (see Figure 5.2). The amounts of connective tissue and smooth muscle vary depending on the vessel diameter, but the endothelium is always present. Capillaries are fine vessels


72 that are only 10 15 m in diameter and consist merely of endothelial cells and a very fine layer of basal lamina. Apart from cartilage, every cell lies within a few m of a capillary, thus mak ing endothelial cells vital for the integrity of almost every tissue of the body. Figure 5 2 Diagram of a small artery in cross section. From [ 100 ] Additionally, endothelial cells have mechanoreceptors that respond to shear stress due to blood flow over their surfaces and by signaling this information to the surrounding cells they enable the vessels to adapt their diameter and wall thickness to suit the blood flow [ 100 ] Endothelial cells also respond to neuronal signaling related to blood vessel dilation by releasing NO gas, to make smooth muscle relax in the vessel [ 100 ] The former phenomena of vasoconstriction and vasodilation, mediated by the endothelial cells, are the processes that regulate blood pressure [ 100 ]


73 Endothelial cells reproduce by duplication of existing endothelial cells. If a part of the vessel wall is damaged and denuded from endothelial cells, neighboring endothelial cells proliferate and migrate to the exposed area [ 100 ] Endothelial cells not only repair existing vessels but also create new vessels (angiogenesis). Endothelial cells line the entire vascular system, from the heart to the smallest capillary, thus a material that is biocompatible with them and also displays low t hrombogenicity would provide an extra advantage for the development of cardiovascular implantable applications (e.g., sensing devices and vascular grafts). 5.2 Cell P roliferation in V itro : E xperimental P rotocol Cells interacting with a biocompatible surface will increase their attachment area and multiply keeping their normal morphology, thus having greater viability than cells in contact with cytotoxic surfaces [ 1 ] [ 37 ] [ 102 ] [ 74 ] It has been proven that 3C SiC has excellent biocompatible properties, however it has also been shown in [ 11 ] that some cells prefer silicon, so in the particular case of hemocompatibility of single cryst al 3C SiC versus silicon, it is necessary to evaluate cell viability and morphology on the substrates of interest In order to assess the biocompatibility in vitro of 3C SiC the methods outlined in the standard ISO 10993 1 [ 2 ] were used including fluorescent microscopy and measurements of cell proliferation. Cell proliferation was measured using yellow MTT (3 (4,5 Dimeth ylthiazol 2 yl) 2,5 diphenyltetrazolium bromide a tetrazole) ) assays [ 9 ] [ 11 ] [ 103 ] Yellow MTT solution is reduced to purple formazan in the mitochondria of living cells. This reduction


74 takes place only when mitochondrial reductase enzymes are active, and therefore conversion can be directly related to the number of viable (living) cells [ 104 ] [ 105 ] The absorbance of this colored solutio n can be quantified by measurement at a certain wavelength (usually between 500 and 600 nm) by a spectrophotometer [ 103 ] The higher the cell proliferation the higher the amount of synthesized forma zan and that translates to an increase d absorbance value The MTT method of cell viability determination is most useful when cultures are prepared in multiwell plates so that statistical anal ysis can be performed to ensure that the data is significant The samples used in these experiments were low roughness single crystal 3C SiC films grown on Si (100) substrates and then diced in to 8mm by 10 mm die The samples were selected from those of the platelet adhesion and morphology experiments described in (§3.2) having flat surfaces with an average roughness of [0.25 + 0.11] nm that was estimated from AFM measurements taken in taping mode It has been r eported that cell proliferation for cells seeded on samples that have been repeatedly cleaned with piranha is reduced over time [ 106 ] It is likely that repeated piranha immersions may decrease the carbon concentration in the near surface region of the SiC samples altering their surface wettability and biocompatibility Piranha cleans of a specific sample should be limited to a maximum of five as a precaution for cell proliferation [ 106 ] As the samples were used in several experiments that involved piranha cleaning, they were hydrogen etched at 150 Torr p ressure flowing 3 SLM of Pd (Palladium membrane) purified hydrogen, for 30 minutes at 1200 C. in a hot wall CVD reactor specifically designed for 3C SiC growth and processing which is described in [ 107 ] as the optimum


75 method for 3C SiC Hydrogen etching 3C SiC and Si die semiconductor samples where cleaned following the protocol described in (§3.2) prior to any cell proliferation experiment. A normal human microvascular endothelial cell (HMVEC) line, cryopreserved from adult dermis and capable of at least 16 population doublings, was obtained from (Cat. C 011 5C) [ 108 ] The cells were cultured in Corning Cellbind 75 mm 2 culture flas ks from Fisher scientific (Cat. 07 202 515) that were Cat. S 006 100) containing gelatin at 0.1% [ 109 ] Cat. M 131 500) sup plemented with Microvascular Growth Supplement (M VGS) S 005 25) [ 109 ] containing fetal bovine serum (5% v/v final concentration), hydrocortisone, recombinant human fibroblast growt h factor, heparin, recombinant human epidermal growth factor, and dibutyryl cyclic AMPand, and 1% penicillin streptomycin. The cells cultures were incubated in a 37 C, 5% CO 2 /95% air, and humidified cell culture incubator. When the cul tures reached 80% 90% confluence they were collected th r ough trypsination (using Trypsin/EDTA solution and Trypsin neutralizer), centrifuging at 180g for 7 minutes and resuspending the cell pellet in supplemented Media 131. The density of the cell suspension was later d etermined using a hemocytometer. The cells were seeded then on multiwell plates (Fisher Cat. 08 772 4G), with flat bottom and enhanced reproductivity due to a stable modification of the polystyrene surface that allows the proliferation of cell lines


76 that grow poorly on standard tissue culture surfaces, and used as controls. Six wells were pre treated with attachment factor prior to cell seeding. Six 3C SiC/Si(100) die and six Si(100) die were used per experiment and eac h experiment was repeated 3 times for each of the samples (i.e. three repetitions and 6 sample replicates per experiment) The samples were seeded with HMVEC after sample cleaning and placement in an ultra low attachment 24 multiwell plate from Corning (Fisher Cat. 07 200 602), in order to prevent cell migration and growth on the well surface in case this was more biocompatible than the semiconductor samples. The area of the wells used in these exper iments was 200mm 2 Figure 5.3 presents a schematic of the sample positioning in the multiwell plates. Figure 5 3 Schematic representation of sample position for cell plating experiments Die size 8mm x 10mm.


77 The cells were seeded on the samples at a density of 15 x10 3 cells/cm 2 (determined experimentally to avoid cell confluence before 168 hours) and the volumes of cell suspension deposited on the samples scaled according to the respect ive plating areas (200 mm 2 for the plates versus 80 mm 2 for the semiconductor die), as described in [ 110 ] followed with the addition of 2 ml of supplemented Medium 131, that was changed on a daily basi s. After seeding, all the plated cells were incubated at 37 C for 72, 120 and 168 hours in an air atmosphere containing 5% CO 2 and 95% relative humidity Next MTT assays measurements were performed, using the BioTek Synergy 2 SLFA plate reader operating at a wavelength of 540 nm, following the protocols in [ 34 ] [ 106 ] [ 103 ] In order to reduce sampling error in the absorbance measurement, 3 drawings of diluted formazan were obtained per sample (replicates) and then averaged. T he measurements were normalized with respect to the average of the treated control wells and expressed as sampling distribution of the mean ( ) standard error of the mean The collected data was compared using ANOVA analysis and paired T test with B onferroni adjustment, in order to determine significant differences between the samples (treated wells, non treated wells, 3C SiC and Si) distributions. The attached H (Green CMFDA from Invitrogen, Cat. C2925) at a working conce n tration of 0.5 M [ 111 ] This probe is retained in living cells after passing freely through the cell membrane and being transformed into cell impermeant fluorescent dye. Living cell staining allowed for cell proliferation, morphology and adhesion quality assessment. Fluorescence microscopy was performed using a Leica DM2000 compound microscope


78 for the control wells and a Leica DM IL inverted microscope for the wafer die. T he results of the MTT assays and f luorescence microscopy are discussed in section (§ 5.3 ). 5.3 MTT Assay and Fluorescent Microscopy R esults Figure 5.4 displays the statistical analysis of the MTT readings after 72 hours. No statistical differences were observed among the different substrates and this was confirmed after ANOVA analysis, which yield ed a large p value ( p =0.598). This result was consistent with the observed degree of cell proliferation on the different substrates, obta ined by fluorescence m icroscopy ( see Figure 5.5 ) The adherent cells showed similar morphology, and good adhesion indicating no preferential adhesion to any particular substrate. Figure 5 4 (Top) Histogram displaying the normalized results ( respect to the treated wells) of the MTT assay after 72 hours. (Bottom) ANOVA table of the measured data


79 Treated Well Non Treated Well 3C SiC/Si(100) Si(100) 20x Figure 5 5 Fluorescent micrographs of HMVEC proliferation on various substrates (treated well, non treated well, 3C SiC and Si) after 72 hours. Previous research on mammalian cell adhesion to SiC has shown that significant differences in cell proli feration are typically observed after increasing the culturing time as long as confluence is not reached [ 106 ] [ 11 ] Following the same rationale HMVEC were c ultured on the different substrates for longer times (i.e. 120 and 168 hours) in order to determine if the cells had a different proliferation rate with time on any particular substrate. Figure 5 6 (Top) Histogram displaying the normalized results ( respect to the treated wells) of the MTT assay after 120 hours. (Bottom) ANOVA table of the measured data


80 Treated Well Non Treated Well 3C SiC/Si(100) Si(100) 20x Figure 5 7 Fluorescent micrographs of HMVEC proliferation on various substrates (treated well, non treated well, 3C SiC and Si) after 120 hours. Figures 5.6 and 5.7 show the statistical analysis of the MTT assays and the fluorescent micrographs for adherent cells after 120 hours of seeding. In this case a higher proliferation was observed on silicon and the result was statistically confirmed by the paired test. The wells treated with attachment factor, the non treated well and 3C SiC substrates had no statistical differences for the cell viability measurement, however Si displayed an increased proliferation that was statically different from all the other samples. After well treatment it was expected to have an increased pr oliferation on these well s than on the non treated wells. H owever as the culture insert used for the controls have a high reprodu ctive nature it actually does not make any difference to use the attachment factor. This is completely different from culturing the cells in conventional culture flasks were if attachmen t factor is not used only a small percentage of the cell suspension is attached to flask culture surface.


81 Figure 5 8 (Top) Histogram displaying the normalized results ( respect to the treated wells) of the MTT assay after 168 hours. (Bottom) ANOVA table of the measured data Treated Well Non Treated Well 3C SiC/Si(100) Si(100) 20x Figure 5 9 Fluorescent micrographs of HMVEC proliferation on various substrates (treated well, non treated well, 3C SiC and Si) after 168 hours. The same trend was observed again after 168 hours (see Figures 5.8 and 5.9). A higher c ell proliferation was observed o n Si, which was as high as 60% more than the proliferation observed for the other cultures, and statistically different from the other substrates (which display no statistical differences among them ). No morphological diffe rences were observed among the substrates. Healthy elongated endothelial cells proliferate on the substrates creating mosaic like patterns that had good


82 quality of adhesion (flattened cells) and no morphological changes as the culturing time increased. 3C SiC displayed consistently good biocompatibility that was stati sticall y comparable to high proliferation culture wells. In [ 11 ] it was reported that certain cell lines can induce physical damage on the substrate surfaces. The Si a nd 3C SiC surface morphology were evaluated before and after culturing (after cleaning) and no differences were observed in surface morphology before and after cell seeding and removal 5.4 Discussion After the excellent platelet adhesion and morphology assessment results that were obtained for 3C SiC, it was very compelling to evaluate its compatibility with the cells that line the cardiovascular system (en dothelial cells). An adult human dermis microvascular cell line was selected and cell proliferation was evaluated on 3C SiC and Si, the leading material for biomedical sensing applications. After statistical analysis of the MTT assays and morphological eva luation of cell cultures it was found that 3C SiC had an excellent biocompatibility with this cell line, comparable with the culture inserts treated with attachment factor or equivalent high reproductivity culture wells. However Si displayed a superb proli feration, about 60% above the culture wells and 3C SiC, without affecting cell morphology (cell were flattened, elongated and well spread on the surface), which means that this high proliferation was not the result of cell mutation or accelerated prolifera tion, but rather the effect of enhanced cell and cell adhesion protein attachment.


83 These results may induce us to think of Si as a superior material with biocompatible characteristic s, which for the case of HMVEC appears to be true, but it does not necessarily mean that the material is more hemocompatible. When interpreting these results one must think of the final application, which in this case is the future development of cardiovascular devices If Si is coated with a lining of endothelial cells, and implanted in the human body, it is likely that its surface is denuded from cell s at some point, exposing a highly thrombogenic surface to the blood stream activating therefore the clotting cascade and presenting an elevated risk or material debris co ntaminating the blood stream due to Si 's lack of mechanical resilience. On the other hand, the reason why Si outperformed 3C SiC is most likely related to the high attachment of adhesion proteins to its surface due to the effect of its hydrophobicity and p rotein layer film interfacial tension [ 84 ] [ 112 ] [ 113 ] e specially Von Willebrand's Factor, which is produced constitutively by the endothelium and is present in the blood plasma [ 100 ] It has been reported in [ 11 4 ] that substrates coated with Von Willebrand's Factor have higher platelet adhesion and aggregation as this factor is responsible for platelet tethering and translocation Therefore it must be considered that protein adhesion adsorption onto surfaces requires a compromise, in this particular case the same reason that enhances cell proliferation on Si substrates is the same reason that makes it perform poorly when in contact with plasma proteins and platelets.


84 Chapter. 6: Conclusions and Future W ork 6.1 Conclusions: 3C SiC V ersus S i T o date, few crystalline semiconductor materials have been assessed for their biocompatibility. Crystalline semiconductor materials are of special interest for the development of biomedical sens ors that can be in intimate contact with the human body Therefore biocompatible sensing materials are critical to the development of implantable biomedical devices and this dissertation was dedicated to evaluating the biocompatibility of such materials for use in the cardiovascular system. The develop ment of implantable sensors that are placed in the blood stream has faced many challenges dominated by the hemocompatibility of the materials that can be used for their fabrication. T he formation of thrombus on the surface of the semicon ductors is of special concern and is a response to the activation of the clotting cascade, and the subsequen t embolization of the thrombus. Many polymeric coatings have been developed or modified in order to reduce the thrombogenicity of engineered materia ls, but these coatings have all degraded over time once they are placed in the human body and thus make long term cardiovascular implants very difficult to achieve The former states the necessity of a semiconductor material that is biocompatible, hemocompa tible and chemically and mechanically resilient, as the material need s to be in contact with the harsh environment of the bod y (i.e., bodily fluids )


85 Two materials were selected for this research in order to compare their hemocompatibility, namely Si and 3 C SiC. Silicon is a material traditionally used for the development of MEMS devices, and is the main semiconductor used for biomedical applications and is known to be biocompatible with certain cell lines. 3C SiC is a chemical ly and mechanically resilient wide band gap semiconductor, very suitable for harsh environments, that has been proven to be biocompatible with several cell lines, but whose hemocompatibility in the single crystal line state has not yet been reported. To the b est of our knowledge this is the first research that directly compares crystalline Si and 3C SiC for their hemocompatibility In order to assess the hemocompatibility of the se semiconductor materials three different aspects involved in thrombosis were used to design the experiments ( see §2.3) : 1) the platelet adhesion and morphology assessment, 2) the protein adsorption perspective and 3) the human microvascular endothelial cell proliferation perspective. Pla telet adhesion and morphology were assessed a fter the exposure of platelet rich plasma ( PRP ) to di fferent SiC polytypes, two different crystal orientations and as a function of surface roughness and the results were contrasted to those of Si(100) wafer die (§3.2) In each case single crystal 3C SiC grown on Si(100) substrates had the best performance, with low particle adhesion counts and small area adhered platelets that also showed high platelet circularity. Si, 4H SiC and 6H SiC displayed high thrombogenicity and were therefore determined to be no n suitable for cardiovascular applications A second experiment was designed (§4.3) to monitor protein adsorption onto the surface of 3C SiC, nano crystalline diamond ( NCD ) and Si coated quartz crystal sensors


86 using QCM D technology Unfortunately the NCD c oatings peeled off and the SiC sensor coating was 4H SiC instead of 3C SiC. However from this experiment it could be confirmed that protei n adsorption to Si and 4H SiC are statistically similar, which confirms the platelet adhesion and morphology results o bserved o n those substrates during the first experiment reported here (§3.2) It was also found that the viscosity of the protein layers adsorbed to the hard substrates wa s smaller than for the proteins adsorbed onto polymeric film coatings. Next, proliferation of human microvascular endothelial cells ( HMVEC ) on 3C SiC substrates was directly compared to Si substrates (§5.2). Cell viability was quanti fied using MTT assays (§5.3) and cell morphology inspected using fluorescence microscopy (§5.4) Aft er statistical analysis of the MTT assays and morphological evaluation of cell cultures it was found that 3C SiC had an excellent biocompatibility with this cell line, comparable with the culture inserts treated with attachment factor or equivalent high re productivity culture wells. However Si displayed higher proliferation than 3C SiC and appears to be more biocompatible with HMVEC Cells proliferate on both substrates without altered cell morphology, which is a good indication of non degenerative interac tions at the cell material interface (good biocompatibility ) [ 115 ] It was concluded that even though Si has a higher biocompatibility for microvascular endothelial cells, this does not necessarily impl y hemocompatibility, as the cell proliferation was assessed in vitro and in the absence of blood interactions. If a material that displays high thrombogenicity also displ ays good HMVEC compatibility one must


87 consider the implications of this material being coated with endothelial cells and then being placed in the blood stream. Under those conditions, i f the material is denuded from endothelial cells then the platelet attachment and activation response will be f aster than the cell's capability to cover the surface [ 116 ] therefore the thrombogenic nature of the base material ( as indicated by platele t adhesion assessment ) will negate the positive endothelial cell biocompatibility result and makes Si non suitable for cardiovascular applications in the long term. The reason why Si outperformed 3C SiC in cell proliferation assessment is most likely related to the high attachment of adhesion proteins to its surface. Therefore a proper balance of the amoun t and type of proteins adsorbed to surfaces is required to guarantee hemocompatibility. 3C SiC demonstrated in vitro characteristics that are highly desirable for cardiovascular applications including low thrombogenicity and good compatibility with endothelial cells. 3C SiC has a better trade off, between cell proliferation and platelet adhesion in vitro than silicon The low thrombo genicity of 3C SiC reduces the chances of catastrophic failure of a device due to thrombi embolization and its resilience reduces the chances of material debris falling into the blood stream 6.2 Future W ork The experiments performed in this research provided insight into the compatibility of 3C SiC and silicon with blood H owever in order to develop implantable sensing devices


88 and two aspects should be explored more on detail: the behavior of the material in vivo and the interaction of the material with plasma proteins. However, in order to address hemocompatibility in vivo hemocompatibility must be first studied under blood flow conditions. Additionally, the study of surface protein interactions is key to truly understand ing the mechanisms behind the hemocompatibility of a material and to functionalize surfaces. Surface functionalization is perhaps the most challenging and the last step required to develop an implantable sensing device. The e nzyme linked immunos orbent assay also called ELISA is a technique used to detect the presence of an antibody or an antigen in a sample This technique can be used in vitro in orde r to detect and quantize platelet activation b y measuring the levels of relevant platelet metabolic products [ 117 ] (e.g. alpha granule components such as beta thromboglobulin and Platelet Factor 4 ) that are specific to platelet release or the presence of the soluble form of the adhesion molecule P selectin [ 118 ] after contact with the semiconductor material surfaces. Raised levels of soluble P selectin are found in a variety of thrombotic disorders and its levels do not seem to be influ enced by the various anticoagulants and different methods of plasma preparation [ 68 ] [ 119 ] making it an ideal candidate for further evaluation of the thrombogenicity of blood contacting materials. These types of measurements will provide additional information regarding platelet activation in the presence of material blood plasma interactions. In order to detect activated platelets under flow conditions the surface expression of P selectin can be performed using fluores cence activated flow cytometry [ 120 ] T his

PAGE 100

89 technique would also involve the detection of antigens on the membrane of activated platelets ( e.g., GP IIb IIIa ), platelet surface bound fibrinogen and it would additionally allow the detection platelet changes in shape of activated circulating platelets in contact with the materials [ 121 ] Future work regarding the coating of quartz crystal sensors with 3C SiC single crystal films would allow the study of fibrinogen and GPIIb IIIa onto this substrate using QCM D If this were possible, the QCM D experiment could be extended to the exposure o f flowing platelets to the layer of pre adsorbed proteins, and quantize the degree of platelet adhesion and activation related to protein adsorption. Following the same rationale, 3C SiC could be compared to sensors coated with hemocompatible polymeric fil ms. Finally, an important step in the development of an implantable sensor would be the study of plasma protein adsorption and platelet activation related to the semiconduc tor surface energy Perhaps the most useful step in the determination of the hemocompatibility of any material would be in vivo biocompatibility studies whereby the material is implanted into the blood stream of a living animal and analyzed following the appropriate amount of time. However this last step is very time consuming and expensive and should only be considered once the previous suggested future work research steps have been completed successfully.

PAGE 101

90 References [1] David F. Williams, "On the mechan isms of biocompatibility," Biomaterials vol. 29, pp. 2941 2953, 2008. [2] ISO 10993 Biological evaluation of medical devices Part 1: Evaluation and testing. [3] S. C. Bayliss, L. D. Buckberry, P. J. Harris, and M. Tobin, "Nature of the silicon animal cell interface," Journal of Porous Materials vol. 7, pp. 191 195, 2000. [4] ISO 10993 Biological evaluation of medical devices Part 5: Tests fot in vitro cytotoxic ity. [5] A. D. Michelson, Platelets San Diego: Elsevier Academic Press., 2002. [6] Surmodics, Inc., Diagnosing hemo(in)compatibility, from clinical and device consideratins to test bench assessment, 2004. [7] Surmodics, Inc., Advancements in blood and tissue compatible coatings, 2004. [8] A. Bolz and M. Schaldach, "Artificial heart valves: Improved blood compatibility by PECVD a SiC:H coating," Artificial Organs vol. 14, no. 4, pp. 260 269, 1990. [9] C. Coletti, M. J. Jaroszeski, A. M. Hoff, and S. E. Saddow, "Culture of mammalian cells on single crystal SiC substrates," in Fall Mater. Res. Soc. Meeting Proc 950 Boston, 2006. [10] A. Pallaoro et al., "Study and characterization of different SiC termi nations and their interactions with biomolecules," in XVIII Congresso Nazionale sulla Scienza e tecnologia del vuoto Florence, Italy, 2007. [11] C. L. Frewin et al., "Atomic force microscopy analysis of central nervous system cell morphology on silicon c arbide and diamond substrates," Journal of Molecular Recognition vol. 22, pp. 380 388, 2009. [12] U. Kalnins, A. Erglis, I. Dinne, I. Kumsars, and S. Jegere, "Clinical outcomes of silicon carbide coated stents in patients with coronary artery disease," M ed Sci Monit vol. 8, no. 2, pp. 116 120, February 2002

PAGE 102

91 [13] M. Reyes et al., "Growth of 3C SiC on Si molds for MEMS applications," in Materials Science Forum vol. 527 529, 2006, pp. 307 310. [14] B. Furie and B. C. Furie, "Thrombus formation in vivo," J. Clin. Invest. vol. 115, no. 12, p. 3355 3362, 2005. [15] M. A. Lichtman et al., Williams hematology 7th ed.: Mc Graw Hill Medical, 2006. [16] M. Frost and M. E. Meyerhoff, "In vivo chemical sensor: tackling biocompatibility," Anal Chem vol. 78, no. 21, pp. 7370 7377, November 2006. [17] S. L. Goodman, S. L. Cooper, and R. M. Albrecht, "The effects of adsorbed albumin on platelet spreading," J. Biomater Sci Polym vol. 2, pp. 147 159, 1991. [18] M. Moriau, E. P. Lavenne, J. M. Scheiff, and C. De beys, "The physiological mechanisms of haemostasis," in Blood Platelets .: Hologramme Ed, 1988. [19] C. Mao et al., "In vitro studies of platelet adhesion on UV radiation treated nylon surface," Carbohydr. Polym. vol. 59, pp. 19 25, 2005. [20] L. Jin et al., "The anticoagulant activation of antithrombin by heparin," in Proc. Natl. Acad. Sci. vol. 94, 1997, pp. 14683 14688. [21] T.I. T. Okpalugo, A. A. Ogwu, P. D. Maguire, and J. A. McLaughlin, "In vitro blood compatibility of a C:H:Si and a C:H thin films," Diamond and Related materials vol. 13, pp. 1088 1092, 2004. [22] S. E. Saddow and A. Agarwal, Eds., Advances in Silicon Ca rbide processing and applications Boston, MA: Artech House, Inc., 2004. [23] S. E. Saddow et al., "Effects of substrate surface preparation on chemical vapor deposition growth of 4H SiC epitaxial layers," Journal of electronic materials vol. 30, no. 3, pp. 228 234, 2001. [24] C. A. Zorman and M. Mehregany, The MEMS handbook: Introduction and fundamentals Mohamed Glad el Hak, Ed.: CRC Press LLC., 2005, ch. 15.8. [25] L. Jiang et al., "Fabrication of SiC microelectromechanical systems," J. Vac. Sci. Technol. vol. 21, no. 6, pp. 2998 3001, 2003. [26] X. Li et al., "Micro/Nanoscale mechanical and tribological characterization of SiC for orthopedic applications," J. Biomed. Mater. Res. B: Appl. Biomater. vol. 72, no. 2, pp. 353 361, 2005.

PAGE 103

92 [27] A. H. Maine, S. C. Bayliss, P. Barr, M. Tobin, and L. D. Buckberry, "Biologically interfaced porous silicon devices," Phys. Stat. Sol. (a) vol. 182, no. 1, pp. 505 513, 2000. [28] X. G. Zhang, Electrochemistry of silicon and its oxide New York: Kluwer a cademic/Plenum Publishers, 2001. [29] J. I. Carrrasso and M. M. Faktor, "Oxidation reactions and potentials of germanium and silicon," in The electrochemistry of semiconductors P. J. Holmes, Ed. London: Academic Press, 1962, pp. 205 255. [30] D. Carri et al., "Initial and follow up results of the Tenax coronary stent," J. Intervent Cardiol vol. 14, pp. 1 5, 2001. [31] J. A. Fournier et al., "Initial results and 6 month clinical follow up after implantation of a silicon carbide coated coronary stent," Rev. Esp. Cardiol. vol. 54, pp. 567 572, 2001. [32] A. Ogami et al., "Short term effect of silicon carbide whisker to the rat lung," Industrial health vol. 39, pp. 175 182, 2001. [33] I. Svensson, E. Artursson, P. Leanderson, R. Berglind, and F. Lindgren, "Toxicity in vitro of some silicon carbides and silicon nitrides: whiskers and powders," American journal of industrial medicine vol. 31, pp. 335 343, 1997. [34] C. Coletti et al., "Biocompatibility and wettability of crystalline SiC and Si surfaces," in IEEE EMBS Proceedings 2007, pp. 5849 5852. [35] A. Naji and M. Harmand, "Cytocompatibility of two coating materials, amorphous alumina and silicon carbide, using human differentiated cell cultures," Biomaterials vol. 12, no. 7, pp. 690 694, 1991. [36] R. Yakimova et al., "Surface functionalization and biomedical applications based on SiC," J. Phys. D: Appl. Phys. vol. 40, pp. 6435 6442, October 2007. [37] T. V Kumari, U. Vasudev, A. Kumar, and B Menon, "Cell surface interactions in the study of biocomaptibility," trend. Biomater. Artif. Organs. vol. 15, no. 2, pp. 37 41, 2002. [38] G. Sagvolden, I. Giaever, E. O. Pettersen, and J. Feder, "Cell adhesion force microscopy," in Proc. Natl. Acad. Sci. vol. 96, 1999, pp. 471 476. [39] C. S. Goodman, "Mechanisms and molecules that control growth cone guidance," Annu. Rev. Neurosci vol. 1 9, pp. 341 377, March 1996.

PAGE 104

93 [40] D. F. Williams, The Williams dictionary of biomaterials .: Liverpool University Press, 1999. [41] C. Salvagnini. (2005) Thrombin inhibitors grafting on polyester membranes for the preparation of blood compatible materials. [Online]. [42] S. R. Hanson, "Blood coagulation and blood materials interactions," in Biomaterials science, an introduction to materials in medicine; B. D. Ratner, A. S. Hoffman, and F. J. Schoen, Eds. San Diego, CA: Elsevier Academic Press:, 2004, p. 332. [43] B. D. Ratner, "Blood compatibility a perspective," J. Biomater. Sci. Polym. Ed. vol. 11, pp. 1107 1119, 2000. [44] A. H. Schmaier, "Contact activation: a revision," Thromb. Haemost. vol. 77, pp. 101 107, 1997. [45] J. Hunt, "Foreign body response," in Encyclopedia of biomaterials and biomedical engineering g. L. Bowlin and G. Wnek, Eds.: Informa Healthcare, 2004. [46] M. Szycher and C. P. Sharma, Blood compatible materials and devices: Perspectives towards the 21st century .: CRC, 1990. [47] Y. Ito, L. S. Liu, R. Matsuo, and Y. Imanishi, "Synthesis and nonthrombogenicity of polymer membrane with surface graft polym ers carrying thrombin inhibitor," J. Biomed. Mater. Res. vol. 26, pp. 1065 1080, 1992. [48] A. Kishida et al., "Immobilization of human thrombomodulin onto poly(ether urethane urea) for developing antithrombogenic blood contacting materials," Biomaterial s vol. 15, pp. 848 852, 1994. [49] D. E. Garfin, "Gel electrophoresis of proteins," in Essential cell biology: Cell structure, a practical approach J. Davey and M. Lord, Eds. Oxford: Oxford University Press., 2003, vol. 1, pp. 197 268. [50] U. T. Seyfe rt, V. Biehl, and J. Schenk, In vitro hemocompatibility testing of biomaterials according to the ISO 10993 4, 2002. [51] ISO 10993 Biological evaluation of medical devices Part 4: Selection of tests for interactions with blood, 2002. [52] J. M. Buchanan, P. J. Upman, and R. F. Wallin. (1998) MD&DI. [Online].

PAGE 105

94 [53] U. T. Seyfert, V. Biehl, and J. Schenk, "In vitro hemocomptibility testing of biomaterials according to the ISO 1993 4," Biomolecular engineering vol. 19, pp. 91 96, 2002. [54] N. Nurdin et al., "Hemocompatibility evaluation of DLC and SiC coated surfa ces," European Cells and Material vol. 5, pp. 17 28, 2003. [55] A. J. Rosenbloom et al., "Nanoporous SiC: A Candidate Semi Permeable Material," Biomedical Microdevices vol. 6, no. 4, pp. 261 267, 2004. [56] J. Y. Chen et al., "Antithrombogenic investigation of surface energy and optical bandgap and hemocompatibility mechanism of Ti(Ta+5)O2 thin films," Biomaterials vol. 23, pp. 2545 2552, 2002. [57] G. Kotzar et al., "Evaluation of MEMS materials of constru ction for implantable medical devices.," Biomaterials vol. 23, no. 13, pp. 2737 2750, 2002. [58] B A. Weisenberg and D. L. Mooradian, "Hemocompatibility of materials used in microelectromechanical systems: Platelet adhesion and morphology in vitro," J Bi omed Mater Res. vol. 60, no. 2, pp. 283 291, May 2002. [59] L. A.W. Sanderson, K. Emoto, J. M. Alstine, and J. J. Weimer, "Characterization of Grafted Poly(ethylene glycol) on Si Wafers Using Scanning Probe Microscopy," Journal of Colloid and Interface S cience vol. 207, no. 1, 1998. [60] G. M.B.F. Snellings, S. O. Vansteenkiste, S. I. Corneillie, M. C. Davies, and E. H. Schacht, "Protein adhesion at Poly(ethylene glycol) modified surfaces," Adv. Mater. vol. 12, no. 24, pp. 1959 1962, 2000. [61] Y. Tak ami et al., "Protein adsorption onto ceramic surfaces," J. Biomed. Mater. Res. vol. 40, no. 1, 1998. [62] N. Ali et al., "Human microvascular endothelial cell seeding on Cr DLC thin films for heart valve applications," Journal of Materials Engineering and Performance vol. 15, no. 2, pp. 230 235, 2006. [63] G. Perego et al., "Functionalization of poly (L lactic co epsilon caprolactone): effects of surface modification on endothelial cell proliferation and hemocompatibility [corrected].," J Biomater Sci Polym Ed. vol. 14, no. 10, pp. 1057 1075, 2003. [64] F. J. Jing et al., "Hemocompatibility of lanthanum oxide films fabricated by dual plasma deposition," Thin Solid Films vol. 515, no. 3, pp. 1219 1222, 2006.

PAGE 106

95 [65] E. Ruckenstein, A. Marmur, and S. R Rakower, "Sedimentation and adhesion of platelets onto a horizontal glass surface," Thromb Haemostas vol. 36, p. 334 342, 1976. [66] S. L. Goodman, M. D. Lelah, L. K. Lambrecht, and S. L. Cooper, "In vitro v. ex vivo platelet deposition on polymer surf aces ," Scanning Electron Microsc vol. I, p. 279 290, 1984. [67] H. R. Baumgartner, R. Muggli, T. B. Tschopp, and V. T. Turitto, "Platelet adhesion, release and aggregation in flowing blood: Effects of surface properties and platelet function," Thromb Ha emostas vol. 35, 1976. [68] S. Kamath, A. D. Blann, and G.Y. H. Lip, "Platelet activation: assessment and quantification," European heart journal vol. 22, pp. 1561 1571, 2001. [69] Y. Jiang, B. Rongbing, T. Ling, S. Jian, and L. Sicong, "Blood compatibility of polyurethane surface grafted copolymerization with sulfobetaine monomer," Coll. Surf. B: Biointerf. vol. 36, p. 27 33, 2004. [70] L. F. Brass, L. Zhu, and T. J. Stalker, "Min ding the gaps to promote thrombus growth and stability," J. Clin. Invest. vol. 115, no. 12, p. 3385 3392, 2005. [71] J. G. McFarland and R. H. Aster, "Evaluation of four methods for platelet compatibility testing," Blood vol. 69, no. 5, pp. 1425 1430, 1987. [72] N. Weber, H. P. Wendel, and J. Kohn, "Formation of viscoelastic protein layers on polymeric surfaces relevant to platelet adhesion," Biomaterials vol. 28, no. 35, pp. 5246 5258, 2007. [73] H. P. Wendel and G. Ziemer, "Coating techniques to im prove the hemocompatibility of artificial devices used for extracorpreal circulation," Eur. J. Cardio thorac. Surg. vol. 16, pp. 342 350, 1999. [74] O Chappey, M P. Wautier, and J. L. Wautier, "Endothelial cells in culture: a model to study in vitro vascular toxicity," Toxic. in Vitro vol. 9, no. 4, pp. 411 419, 1995. [75] B. J. Bain, A beginners guide to blood cells 2nd ed. Massachusetts: Blackwe ll Publishing Ltd, 2004. [76] M. M. Frojmovic and J. G. Milton, "Human platelet size, shape, and related functions in health and disease," Physiological reviews vol. 62, no. 1, pp. 185 261, January 1982.

PAGE 107

96 [77] (2003, December) The journal of Biologica l Chemistry. [Online]. expansion [78] K. Park, F. W. Mao, and H. Park, "Morphological characterization of surface induced platelet activation," Biomaterials vol. 11, pp. 24 31, 1989. [79] Image J. [Online]. [80] S. Levy Toledano, J. Maclouf, P. Bryon, and E. H. Savariau, "Human platelet activation in the absence of aggregation: A calcium dependent phenomenon independent of thromboxane formation," Blood vol. 59, no. 5, pp. 1078 1085, May 1982. [81] H. Mani, B. Luxembourg, C. Klffling, M. Erbe, and E. Lindhoff Last, "Use of native or platelet count adjusted platelet rich plasma f or platelet aggregation measurements," J Clin Pathol vol. 58, p. 747 750, 2005. [82] S. J. Hitchcock, N. T. Carroll, and M. G. Nicholas, "Some effects of substrate roughness on wettability," Journal of Materials Science vol. 16, no. 3, pp. 714 732, Marc h 1981. [83] J. Israelachvili, Intermolecular and surface forces London: Academic Press, 1992. [84] R. I. Freshney, Culture of animal cells: a manual of basic technique 5th ed.: John Wiley & Sons, Inc., 2005. [85] L. J. Yao and K. H. Mayo, "Interactions of integrin GPIIb/IIIa derived peptides with fibrinogen investigated by NMR spectroscopy," Biochem. J. vol. 315, pp. 161 170, 1996. [86] Y. Gu, P. Boonstra, A. A. Rijnsburger, J. Haan, and W. van Oeveren, "Cardiopu lmonary bypass circuit treated with surface modifying additives: a clinical evaluation of blood compatibility," Ann. Thorac. Surg. vol. 65, pp. 1342 1347, 1998. [87] Gladstone. (2009, October) Integrilin: Basic Research Translated into Therapy. [Online]. 1396 [88] M. S. Lord et al., "The effect of silica nanoparticulate coatings on serum protein adsorption and cellular response," Biomaterials vol. 27, pp. 4856 4862, 2006. [89] KSV. (2004, October) KSV QCM techinical note. Electronic document.

PAGE 108

97 [90] N. Weber, A. Pesnell, D. Bolikal, J. Zeltinger, and J. Kohn, "Viscoelastic properties of fibrinogen adsorbed to the surface of biomaterials used in blood contacting medical devices," Langmuir vol. 23, pp. 3298 3304, 2007. [91] Q sense. (2009) Q snese QCM D technology. [Online]. http://www.q [92] M. Rodahl and B. A. Kasemo, "A simple setup to simultaneously measure the resonant frequency and the absolute dissipation factor of a quartz crys tal microbalance," Rev. Sci. Intrum. vol. 67, no. 9, pp. 3238 3241, 1996. [93] K. A. Marx, "Quartz crystal microbalance: a useful tool for studying thin polymer films and complex biomolecular systems at the solution surface interface ," Biomacromolecules vol. 4, no. 5, pp. 1099 1120, 2003. [94] B. Osterud, "Activation pathways of the coagulation system in normal haemostasis," Scan. J. Haematol vol. 32, pp. 337 345, 1884. [95] W. B. Tsai, J. M. Grunkemeier, C. D. McFarland, and T. A. Horbertt, "Platelet adhesion to polystyrene based surfaces preadsorbed with plasmas selectively depleted in fibrinogen, fibronectin, or von Willebrand's factor," J. Biomater. Res. vol. 60, no. 3, p p. 348 359, 2002. [96] Q Sense. Hemocompatibility of hydroxyapatite (Application note QS405 12 1). electronic document. [97] J. W. Smith, Z. M. Ruggeri, T. J. Kunicki, and D. A. Cheresh, "Interaction of integrins alpha v beta 3 and glycoprotein IIb IIIa with fibrinogen. Differential peptide recognition accounts for distinct binding sites," J. Biol. Chem. vol. 265, no. 21, pp. 1 2267 12271, 1990. [98] FCoE BITT. (2009) USF FCoE BITT. [Online]. [99] N. Weber, H. P. Wendel, and J. Kohn, "Formation of viscoelastic protein layers on polymeric surfaces relevant to platel et adhesion," Journal of Biomedical Materials Research Part A vol. 72A, no. 4, pp. 420 427, 2005. [100] B. Alberts et al., Molecular biology of the cell 4th ed. New York: Garland Science, 2002. [101] J. McGeachie. (1998, February) Blue histology mor e about endothelial cells. [Online].

PAGE 109

98 [102] R. G. Richards, "The effect of surface roughness on fibroblast adhesion in vitro," Injury vol. 27, no. Supl 3, pp. S C38 S C48, 1996. [103] T. Mosmann, "Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays," Journal of immunological methods vol. 65, no. 1, pp. 55 63, 1983. [104] D. M. Spinner, "MTT Growth Assays in Ovarian Cancer," i n Ovarian Cancer: Methods and Protocols M. S. Bartlett, Ed.: Humana Press, Inc, pp. 175 177. [105] S. Ahmadian, J. Barar, A. A. Saei, M.A. A. Fakhree, and Y Omidi. (2009) Cellular Toxicity of Nanogenomedicine in MCF 7 Cell Line: MTT assay.JoVE. 26.. [Onl ine]. [106] C. Coletti, Silicon carbide biocomaptibility, surface control and electronic cellular interaction for biosensing applications, 2007, Ph.D .E.E. dissertation. [107] C. L. Frewin, C. Coletti, C. Riedl, U. Starke, and S. E. Saddow, "A comprehensive study of hydrogen etching on the major SiC polytypes and crystal orientations," Materials science forum vol. 615 17, pp. 589 592, 2009. [108] Cascade Biologics. Inc. Human Microvascular Endothelial Cells (adult dermis). Document. [Online]. [109] Cascade Biolo gics. Inc. Medium 131, MVGS and AF: Instructions for storage and use. Document. [110] C. C. Perry and T. Keeling Tucker, "Aspects of the bioinorganic chemistry of silicon in conjunction with the biometals calcium, iron and aluminium," J. Inorg. Biochem. vol. 69, no. 181, 1998. [111] Term Tracing of Living Cells. Document. [112] Barry Arkles. (2006, October) Hydrophobicity, hydrophilicity and silanes. Paint & Coatings Industry Magazine. [113] L. Mirenghi, P. A. Ramires, R. E. Pentassuglia, P. Rotolo, and A. Romito, "Growth of human endothelial cells on plasma treated polyethyleneterephthalate surfaces," Journal of material science: materials in medicine vol. 11, pp. 327 331, 2000. [114] S. Ku lkarni et al., "A revised model of platelet aggregation," The journal of clinical investigation vol. 105, no. 6, pp. 783 791, March 2000.

PAGE 110

99 [115] A. J. Garca, "Intefaces to control cell biomaterial adhesive interactions," Adv. Polym. Sci. vol. 203, pp. 1 71 190, 2006. [116] P. Zilla, M. Deutsch, and J. Meinhart, ""Endothelial cell transplantation," Semin. Thromb. Hemostais. vol. 12, pp. 52 63, 1999. [117] M. Fisher, P. H. Levine, and A. L. Fullerton, "Marker proteins of platelet activation in patients w ith cerebrovascular disease," Arch. Neurol. vol. 39, pp. 692 695, 1982. [118] A.D. Blann and G.Y.H. Lip, "Hypothesis: is soluble P selectin a new marker of platelet activation?," Atherosclerosis vol. 128, pp. 135 138, 1997. [119] B.H. Chong et al., "Plasma P selectin is increased in thrombotic consumptive platelet disorders," Blood vol. 83, pp. 1535 1541, 1994. [120] L. Corash, "Measurement of platelet activation by fluorescence activated flow cytometry," Blood Cells vol. 16, pp 97 106, 1990. [121] A.H. Lazarus, J.F. Wright, V. Blanchette, and J. Freedman, "Analysis of platelets by flow cytometry," Transfus Sci vol. 16, pp. 353 361, 1995.

PAGE 111

100 About the A uthor Norelli Schettini was born on December 7 th 19 82 in Barranquilla ( Atlntico), Colombia In 2005, she earned her bachelor's degree in Electrical Engineering from the Universidad del Norte (Barranquilla, Colombia). She graduated with honors and received the scientific merit award for her research work on spee ch recognition in cotton top tamarins ( saguinus oedipus ), an endemic endangered species from the north coast of Colombia. In 2005 she was selected as a scholar by Universidad del Norte, to become part of the faculty after acquiring her Ph.D. degree at Uni versity of South Florida (USF) in Tampa, FL (USA), as part of a special program to improve education and research in South America. During her Ph.D. studies at USF, she was awarded the FCoE BITT (Florida Center of Excellence, Biomolecular Identification an d Targeted Therapeutics) fellowship to conduct her research on hemocompatibility of Single crystal 3C SiC.