\documentclass[a4paper,10pt]{report}
\usepackage[T1]{fontenc}
\usepackage[latin1]{inputenc}
\usepackage[english]{babel}
\usepackage{amsfonts}
\usepackage{graphicx}
\usepackage{float}
\usepackage{url}												% per creare collegamenti attivi a siti web
											

\DeclareGraphicsExtensions{.png,.pdf,.jpg,.gif}

\begin{document}
\begin{titlepage}
\author{Katia Donato}
\title{Characterization of a proton and an alpha beam for radiobiology: Evaluation of broad beam performances from a 2 MV tandetron.}
\maketitle
\centering
A.A. 2006/2007
\\Leonardo report
\thispagestyle{empty}
\end{titlepage}

\tableofcontents
\thispagestyle{empty}
\newpage
\section{Introduction}
During the period from 17th of April to 21st of July spent in Belgium, I worked with IBA and in a lab involved in numerous collaboration with IBA. In
this Lab (LARN) I have worked with a particle accelerator ALTAIS (Accélérateur Linéaire pour l'Analyse et l'Implantation des Solides). It is a 2 MV
terminal voltage tandetron currently used for material synthesis (ion implantation) and material analysis (RBS, NRA, PIXE). Their intention is to use
this accelerator, in collaboration with IBA, to perform basic studies about particle-radiobiology, IBA being heavily involved in particle therapy.
The aim of this work was to characterize a proton and an alpha beam at the end of a beam line that will be used for radiobiology experiments.The particularity is to produce and monitor a beam homogeneous over 0.5 cm$^{2}$ by using only the defocusing capabilities of the installation, and without
scanning the beam.
The first part consisted in studying literature by making a literature work on accelerators applied to cells'irradiation (Chapter 1).In particular this study was focussed on understanding how to build up a beam and to monitor it while irradiating a colony of cells. 
The second part is dedicated to the beam monitoring and beam homogeneity studies. Previously X-ray emissions for a thin layer induced from particle transit was characterized as a beam monitroring device. The layers were synthetized in the laboratory by means of a PVD set up. First, X-Ray production yield were simulated simultaneously to understand if it was possible to detect and follow beam variations. Second, a scintillator was evaluated as beam profile too and simulations were made to know the energy loss in the scintillator to decide the best working condition, in particular the thickness to be used in the experiments. The homogeneity and stability check of the two beam were also performed. In the Chapter III, I present the results of the few hours I spent working in preparing the colonies and in irradiating them with a particle beam and a $\beta^{-}$ emitter.
Finally, I conclude by sying that our measurements showed that this installation can be used to perform broad beam irradiation over surfaces larger than 0.5 cm$^{2}$. For proton, no major modifications are requested apart the engineering of a beam monitoring device based on plastic scintillation. For
$\alpha$, the ion-source regulation is not stable enough to achieve a beam current stability. Work is on going to adapt the Li thermal exchange canal for better temperature regulation. For heavier ions, a special exit section has to be designed. One paper was written at the end of this work.
\newpage
\chapter {State of the art of accelerators applied to radiobiology}
\section{Introduction}
The study on the interaction of charged particles with living matter is important for medical application. Infact it can be used both for therapy and
protection. On one hand it is interesting to see the damage induced by natural radioactivity (for example radon) and on the other one to apply particle
beams to produce a damage in unhealthy tissues or cells.
Today there are a lot of experiments concerning radiotherapy and radioprotection. In particular it is useful to study how ion beams interact with cells in terms of damage and dose released to them.
It is known that the damage induced in a tissue has a dependence on the energy, on the mass of the ion and on the type of tissue irradiated. Linear energy transfer (LET) increase with particle mass, this means that heavier ions release more energy in a medium that lighter ones. The biological effect of high LET radiations is in general higher than those of low LET radiations with the same energy. The RBE (radiobiological effectivness) increase also with the energy of the ion beam, so carbon ion of 80 MeV induce more damage than 10 MeV one.
Ion beams allow the irradiation of tumor volumes by higher doses while sparing with healthy tissue and increase the penetration depth forming a very sharp peak near the stopping region (Bragg peak).
An experiment to evaluate the biological damage induced by ion beams is the irradiation of a culture of cells. The use of a particle accelerator allows to obtain a large range of energy and also a wide range of ion accelerated. There are different ways to deliver an experiment of radiobiology and this depends on the type of beam used, the type of damage induced and also the set up.
The tipical set up consists of an accelerator that accelerate different ions at different energies, a detector to check the beam before irradiating the sample (homogeneity, position, size) and after the sample to know the dose delivered to the cells, a positioning system for the sample to be irradiated.
There is also an important part of preparation of the sample that can be deliver in different ways according to the study and to the type of cells under exam.
In this paper will be discussed some ways to deliver experiments on cells focussing on the size of beam used (broad beam or microbeam) and on the irradiation line (vertical or horizontal) reporting the set up and data of various research center and laboratories.

\section{Experiments with broad beam}
\subsection {Gray Cancer Institute}
One way to irradiate cells is the use of a broad beam, so there is no need of a system to focalise the beam to small size. Scattering foils are used to obtain a broad homogeneous beam.
The Gray Cancer Institute has dealt some experiments with ions using a broad beam configuration \cite {report1}. The energetic particles were procuced using a 4 MV Van de Graaff accelerator. The particles available were protons in the energy range 1.0-3.7 MeV, deuterons in the energy range of 0.9-3.4 MeV and helium ions in the energy range 3.4-6.9 MeV. This correspond to volume-averaged LET between 10 and 28 $keV\mu m^{-1}$ for protons, 18-50 $keV\mu m^{-1}$ for deuterons and 59-106 $keV\mu m^{-1}$ for helium ions. 
The samples stay on a motorised sample support that is a 29 cm diameter rotating alluminium platter that sweeps each sample once through radiation field. The speed of rotation (and hence the dose) can be preset for each sample.
The exit window of the beamline is made of 13 $\mu$m or 25 $\mu$m of Kapton. This is necessary for delivering the beam from the accelerator, where there is vacuum, to the sample that, for surviving, must stay in controlled atmosphere. 
A slit that defines the shape of the field is mounted before the exit window; the edges of the slit define a near rectangular field (the edges are not parallel because they are aligned with radii of rotating platter) of about 1 x 26 mm.
To controll the dose delivered to the sample, there are 4 charge collecting plates that are enabled by a position sensor on the wheel that provides a signal to the wheel only when the sample is crossing the field.
A 1 $\mu$m thick gold scattering foil is mounted 1 m from the exit window in the vacuum to obtain an homogeneous beam. The scattering area is reduced by a slit to 1 x 30 mm.
The window slit monitor assembly is positioned near the scattering foil to ensure that only particles scattered by the foil can pass through the exit window.
A parallel-plate chamber is used to measure the dose to the sample position. Particles enter the chamber cavity through the polarizing electrode, which is made of 3 $\mu$m thick aluminazed Mylar. The other electrode is a 13 mm of copper plate. The chamber can be mounted instead of the rotating samples at different position and so it is possible to know the dose delivered to the samples.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.35]{gfx_gci/gci1}
  \caption{Overall view of the sample irradiation apparatus. \label{fig:gci1}}
\end{figure}
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.30]{gfx_gci/gci2}
  \caption{Plan view of the sample irradiation apparatus. \label{fig:gci2}}
\end{figure}
\\
\\
\\
\subsection {Harwell}
The facility available by the Atomic Energy Authority at the Harwell Research Establishment \cite {report2} were a variable energy cyclotron and a Tandem Van de Graaff accelerator. The latter has energy limitation that were resolved by using the cyclotron.
\begin{figure}[!h]
  \centering
  \includegraphics[width= 12 cm]{gfx_harwell/harwell1}
  \caption{ Physical characteristics of the charged particle beams used in mutation and inactivation studies with mammalian cells.\label{fig:harwell1}}
\end{figure}
The tandem accelerator is a 7MV maximum terminal voltage. The can deliver helium ions of 18 MeV. The beam from the main analyser magnet passed through a tantalum collimator and was focused on a quartz plate to 1 or 2 cm in diameter. Then it passed through a gold scattering foil (5 m from the exit window)that produced divergence of the beam and improve the uniformity of the irradiation field and then through further tantalum collimators. The beam emerged from a vacuum exit window that consists of a Melinex foil (4 cm diameter and 10 mg cm$^{-2}$ thickness) coated with a conducting layer of alluminium. The beam was monitored by the secondary emission current produced when it passed through the foil positioned close to a bias electrode (+120 V). The integrated current is proportional to the dose received to each sample. The monitor sistem was calibrated against an ionization chamber.
The cyclotron allowed to achieve higher energies 84 Z$^{2}$/M, depending from the atomic number and the mass. However the set up is similar to those used for Tandem.
Beam current is measured with a Faraday cup.
The irradiation stage is a motorized one that facilitated the simultaneous exposure of 20 specimen dishes. It is placed on a wheel with a drive motor that rotates it.
\subsection {Ganil}
The first irradiation experiments with heavy ions at Ganil (Grand Accelerateur des Ions Lourds) are three delivered in different facilities. The exposures to uranium and argon ions are done at Unilac linear accelerator (GSI, Darmstadt, Germany), one carbon irradiation (200 MeV) is done at SIS Cyclotron (GSI), and another experiment with carbon ions is performed at Ganil accelerator \cite {report3}. The delivered fluences are $10^6/10^8$ particles/$cm^2$.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.415]{gfx/tabella}
  \caption{Dosimetric data for the 4 ions.\label{fig:tabella}}
\end{figure}
They  developed a facility built from the SME exit of their cyclotron \cite {Ganil}. This is the medium energy exit where the energy range is from 3.7 to 13.7 MeV/A.
They also have two other exits: high energy and low energy ion beams. The energy achieved are from 24 to 96 MeV/A for the higher one and from 0.3 to 1 MeV/A for lower one. They can deliver ions from Carbon to Uranium. The intensities reached are 2.10$^{13}$ pps $ ^{12}C^{6+}$ at 95 MeV/A and 7.10$^{12}$ pps for $ ^{64}Ni^{28+}$ at 60 MeV/A. 
They are now studying the damage induced by $ ^{36}Ar$ at 95 MeV/A \cite {Ganil2}. Cells are grown and irradiated in a CR39 track detector. The dose delivered is about 1 Gy.
They are going to start experiments with dose of 2, 4 or 6 Gy using a broad beam (4x40 cm$^{2}$)\cite {Ganil3} and using an ionization chamber to do dosimetry. They simultaneously irradiate the samples for different period of time to study the damage induced by different exposure (dose).

\subsection {GSI}
At GSI there are two facilities: one for low energy beam and another one for high energy beam.
The first (X6) is situated at UNILAC and they can deliver ion beams from protons to uranium with energies inferior to 15 MeV, LET range from 10 to 15000 keV/$\mu$m and fluence from $10^3$ to $10^{10}$ particles/$cm^2$.
The maximum field size is restricted to 50mm. Standard irradiations are performed in 35mm Petri dishes. The samples are positioned in the beam by means of an automatic vacuum grabber system and placed back after irradiation in the magazine filled with cell culture medium; the maximum capacity of a magazine is 20 samples. Other targets can be positioned manually in the beam position.
They use ionisation chamber for dosimetry and a CR39 solid track detector for calibration.
The other facility is for high energy and it can deliver ion beams from protons to uranium, with an energy range from 50 to 2000 MeV and a particle fluence from 1 to $10^{12}$ ions/$cm^2$.
The beam deliver system has a maximum fieldsize of 20 x 20 $cm^2$, an excellent homogeneity for large area irradiation and an on-line monitor to control irradiation. They use a robot arm for positioning the biological samples \cite {gsi}.

\subsection {Bruyères-le-Chatel}
The facility realised in Bruyères-le-Chatel\cite {report4} consists of a 7MV Tandem accelerator for proton and $\alpha$ particle irradiation. The irradiation system allows mono-layer cell irradiation, at atmospheric pressure with large homogeneous beam. To obtain a broad beam it is used a gold scattering foil on which the ion beam is multiple scattered. The number of particles on biological samples per area and time unit is controlled on-line using a NE102 type plastic scintillator for low flux and a Faraday cup for higher flux. A CR39 solid-track detector is used to calibrate both.
They have performed experiments with protons and $\alpha$ particles.
After the acceleration the beam pass through an analyzing magnet to select the charge, then in a collimation system to define the beam and finally it is scattered by a gold foil (5 $\mu$m thick for protons and 10 $\mu$m thick for $\alpha$) to obtain a large irradiating area of at least 1 $cm^{2}$.
The beam is on-line monitored by 5 Faraday cups. The end of the irradiation line is made up of five different exits, two of which are effectively used: the central position, where the biological sample were exposed and the lateral position, where is located a scintillator detector to monitor the dose.
Ion beams are extracted from the vacuum to the atmosphere where there are the samples, through an exit window of Havar (5 $\mu$m thick) for $\alpha$ and one of Kapton (25 $\mu$m) for protons. 
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.60]{gfx_bru/bru1}
  \caption{ Schematic description of the set-up for the adherent cell culture irradiation installed on the 7 MV Tandem accelerator beam line.\label{fig:bru1}}
\end{figure}
The samples were placed on a translatory stage controlled by a computer. The thickness of the sample has to be smaller than the ion range to avoid that they stop in it. The irradiation of the samples is made in 2 different ways: a mono-layer adherent cells on slides without culture medium or mono-layer adherent cells seeded in sterile wells completed with culture medium. In the first it is used a laminar air-flow workstation of 2 x 2 $m^{2}$ area situated 50 cm above the translatory stage to ensure sterile condition. In the second sterile wells are used.
The cells are plated on a 3.5 $\mu$ m thick Mylar foil and then in a well with geometrical cross section of 1.4 $cm^2$. Before the irradiation culture medium is added to the wells that are then close hermetically. Each well could stay out of the incubator just for 30 minutes.
For low particle fluxes (from $10^{4}$ to 5 x $10^{5}$ particles/$cm^2/s$), a 500 $\mu m$ thick NE102-plastic scintillator with a 1.5 $cm^{2}$ associated to a PMT \cite {report4} .
CR39 solid-state nuclear track detectors are used to evaluate the beam uniformity and to check the calibration before and after the exposition.
For high particle fluxes (from $10^6$ to $10^8$ particles/$cm^2/s$) five Faraday cups positioned at 20 cm from the samples were used.
The range of energy used goes from 2.6 to 5 MeV for protons and from 8 to 11.5 MeV for $\alpha$. At the biological sample entrance the energies are from 0.8 to 4.0 MeV for protons and from 3.0 to 8.0 MeV for $\alpha$.
The fluxes are adjusted to deliver 0.5 Gy/s (for low fluences) and 2.5 Gy/s (for high fluences).
The scattering foil are chosed to obtain an homogeneous distribution of beam intensity. As previous said, they decide to use gold 5 $\mu$m thick for protons and 10 $\mu$m thick for $\alpha$.
\subsection {University of Naples} 
The facility for radiobiological irradiation in Naples \cite {report5} consists of a 3MV Tandem accelerator. They irradiate a monolayer of cells with a broad and homogeneous beam of 1 $cm^2$ area. The ion range available in this research centre is from hydrogen to carbon.
The following set up can deliver beryllium, boron and carbon ions at a rate of about 1 $Gy$ $min^{-1}$ but they have also delivered experiments with proton beam.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.50]{gfx_napoli/irradiation}
  \caption{ Dosimetric data for 3 ions irradiation.\label{fig:irradiation}}
\end{figure}
The intensity profile of the beam is measured by a Faraday cup. The beam is collimated with a 2 mm collimator before entering the scattering chamber. In this there is an other collimator (1 mm) and a scattering gold foil that allows to obtain a quite homogeneous beam 1 cm large. Out of the scattering chamber the beam pipe diameter becomes 10 cm to house 4 silicon detectors to monitor the beam during irradiation.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.43]{gfx_napoli/beamline}
  \caption{The beam pipe of Tandem accelerator. \label{fig:beamline}}
\end{figure}
The estraction of the beam from the vacuum to the air is made by a circular window of aluminised Mylar (1.9 $\mu m$ in thickness and 1.3 cm in diameter), that is supported by a grid of 4 wires of tungsten (50 $\mu$m) one horizontal and three vertical to preserve the breaking of it under pressure gradient.
The dosimetric system is formed by the silicon detectors and a CR39 detector of 1.6 mm of thickness. They are both use to take the dose profile of the beam before irradiating the sample. They allow the counting of the particles per $cm^{2}$ (fluence) delivered to the sample and using a program (SRIM) to evaluate the linear energy transfer at the cell mid-plane (the program considers the interaction of charged particles with matter and they stopping power), it can be known the dose to the cells. The CR39 detector check also the spatial uniformity of beam in term of track density.
Beam fluence is recorded with 4 silicon detectors positioned on the exit window. The beam monitor is calibrated using a CR39 detector. The check is made in nine fields of the detector (area of 1.2 x $10^{-4}$) showing a good uniformity.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.40]{gfx_napoli/9field}
  \caption{Nine fields of CR39 plastic detectors with an area of 1.2x10$^{-4}$ cm$^2$.\label{fig:9field}}
\end{figure}
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_napoli/results}
  \caption{LET values at cell mid-plane. \label{fig:results}}
\end{figure}
The preparation of the cells is made putting them, 48 hours before irradiation, in Petri dishes where cells can grow as a monolayer on a Mylar foil (3 $\mu$m thickness). After the irradiation they are detached in trypsine, counted and plated. After 8 days from irradiation it is determined the number of colonies.
\subsection {Infn Legnaro}
The radiobiological facility in Legnaro \cite {report6} consists of a  7 MV CN Van de Graaff accelerator for light ion beams. They installed a single-ion microbeam apparatus at the horizontal broad-beam facility to allow alternatively a single-cell single-ion irradiation as broad-beam irradiation.
This facility delivers beam of protons, deuterons,$ ^{3}He$ and  $ ^{4}He$ ions that have a range from 7 to 150 keV/$\mu$m and it is based on 2 scattering chambers housing gold foils which broad the beam over 1 cm diameter area and reduce its intensity current.
Each chamber contains a tantalum beam collimator (1mm diameter), a beam diffusing gold foil and a silicon surface barrier detector (SSBD) \cite {report24}. A fast tantalum shutter in the first chamber to intercept the beam. The beam line is 150 cm long a 51 mm in diameter and contains three tantalum collimators. The exit window is a 15 mm diameter and 10 $\mu$m aluminized Mylar. A detector is placed outside the beam line to check the beam before irradiation.
The cells are placed in a stainless steel Petri dish. They are grown as a monolayer on a mylar foil (13 mm diameter and 52 $\mu$ thick) and put on the bottom of the dish.
The use of two scattering folis allows to achieve a good stability also at low dose rate. For example they tested a low energy proton beam of 1 nA, broadened over 1 cm$^2$, and they obtained a dose rate of 10$^4$ Gy/min with a single scattering foil and a dose rate of 1 Gy/min with two gold foils. The scatterer is connected to a beam current digitalizer for beam focussig and control. An external micrometric device is used for collimator alignment.
Beam dosimetry is obtained using three SSBD detectors: two to monitor the beam line and another at the sample position. Their useful area is of 1 cm$^2$. The relationship between the dose rate was detemined before each irradiation.
They tested the setup with a 3 MeV proton beam showing good results in beam stability and homogenity. They also characterized using CR39 detector.
They irradiated a cell culture with a beam accelerated to 3 MeV corrisponding to 1.1 MeV on the cells. The dose rate was 1 Gy/min.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.35]{gfx_infn/setup}
  \caption{Schematic view of the proton irradiation facility of INFN Legnaro. \label{fig:setup}}
\end{figure}

For heavy ion broad beam they use the ALPI accelerator facility that is a linear superconducting quarter-wave resonant cavities accelerator with a Tandem accelerator injector. The ions available are $ ^{12}C$  and from $ ^{28}Si$ to $ ^{197}Au$ with an energy of about 20 MV x ion charged state.
Using the last one, they delivered experiments with carbon beams with 94 keV/$\mu$m. Cells were plated on a stainless steel cilinder of 13mm of diameter with a 6 $\mu$mthick mylar base and irradiated with doses ranging from 0.25 to 2.0 Gy.

\subsection {National Institute of Radiological Sciences(HIMAC)}
The facility of the National Institute of Radiological Sciences consists of 2 different set up. Infact they deliver experiments both with broad-beam and with microbeam \cite {report7}. 
The broad beam is produced at the cyclotron of Heavy Ion Medical Accelerator in Chiba (Himac), in particular they are developing a set up for irradiating cells with Carbon ions. They delivered some experiments using  290 MeV/$\mu m^2$ $\ ^{12}C$. The dose-averaged LET of the ion beam is selected to be 100 keV/$\mu$m.
They calculated that when a culture is irradiated by 0.01 Gy and 0.2 Gy of broad-beam, 57.5 \% and 100 \% of the cells could be traversed by at least one particle.
They can accelerate ion beams from He to Argon and reach a maximum energy of 800 MeV/$\mu m$. 
They developed a non-destructive beam-profile monitor using micro-channel plates (MCP) and a resistive anode that can be applied to a detector \cite {report8}. This monitoring system works by charged division to determine the spatial position of an event in gas ions on the detector. 
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_himac/himac1}
  \caption{A micro-channel plate.\label{fig:himac1}}
\end{figure}

\newpage
\section {Experiments with microbeam}
\subsection {Introduction}
A sample can be also irradiated with microbeam that allows to target precise point in the cell. Infact a spatial resolution of a nanometer can be reached by using a microbeam.
Microbeams can be used to deliver low doses of radiation such as that of a single radiation track traversing a cell or a group of cell or to probe subcellular target such as the nucleus and the cytoplasm. It is also possible to study the propagation of information about DNA damage (bystander effect).
Research laboratories have recently discovered the possibility of this type of study and they deliver some experiments on biological culture with ion microbeams.
There are two possible configuration of set up: one using a horizontal beam and the other using a vertical beam.
The main difference consists on the way of positioning cells. In the horizontal configuration the samples are positioned in a vertical position and this implies a limited period of irradiation; infact gravity plays an important role for reaching a stable position and it is mantained just for a little. 
The vertical way allows to position the cells  in a "with gravity" configuration but it's more complicated to obtain a beam like that.
A horizontal line microbeam is used at CENBG (Centre d' Etudes Nucléaire de Bordeaux Gradignan), at the INFN (Istituto Nazionale di Fisica Nucleare) of Legnaro, at SNAKE Superconducting Nanoscope for Applied Nuclear (Kernphysicalische) Experiments in Munich, at GSI in Darmstadt and a nanoprobe facility is under developement at LIPSION of Leipzig. 
A vertical line microbeam is used at Physicalisch Technische Bundesanstalt (PTB) in Germany, at Surrey Ion Beam Centre, at JAERI facility of TIARA, at NIRS facility SPICE (Chiba) and at Columbia University in RARAF.
\subsection {Horizontal line}
\subsubsection {CENBG}
The radiobiological facility at CENBG consists of a 4 MV Van de Graff accelerator that provides a horizontal beam \cite {report9}. The lens configuration is based on a Russian quadruplet that for its simmetry, it allows the use of round-shaped collimators. In this way they can deliver low intensity ion microbeam. In this configuration there are two iridiated platinum circular collimator with anti-scattering shape. The first (5 $\mu$m diameter) limits the intensity of the beam entering the beam line and the second (20 $\mu$m diameter), placed 5.4 m downstream in front of the optics, sets the angular divergence. For proton and $\alpha$ beams up to 3.5 MeV they achieve a lateral resolution of 1 $\mu$m.
The control of particles delivered is made in two different ways. After traversing deflector plates, there is a system to trigger the switching. They designed two transmission detectors. The first is based on a thin plastic scintilator foil (7 $\mu$m thick BC400) coupled to two PMT working in coincidence. They tested the foil by determining the coincidence with an implanted silicon particle detector. The efficiency achieved allows the control of a few thounsands of ions per second. The second is used for $\alpha$ and it consists of a low pressure gas detector (5 mbar, isobutane) equipped with an ultra-thin window.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.40]{gfx_cenbg/cenbg1}
  \caption{Scheme of the gas.\label{fig:cenbg1}}
\end{figure}
The anode is a gold wire 100 $\mu$m diameter positioned 3mm above the ion trajectory. The exit window is constituted by 1 x 1 $mm^2$ $Si_{3}$$N_{4}$ window, 100 $\mu$m in thickness. This window degraded the lateral resolution and they overcome this problem by using a 10 $\mu$m collimator. 
They are developing the irradiation stage equipment. The system is installed on a removable cover of the chamber. The facility is designed to target adherent cells cultured on thin polymer foils in a vertical position. During irradiation the cells are kept in culture medium. They use a fast optical recognition system coupled with a fluorescence microscope and a computer-controlled high sensitivity CCD camera. The aim is to obtain an automaic irradiation procedure with a targeting rate of above 1 cell per second.
The beam resolution and the positioning accurancy is tested by the use of a CR39 detector.

\subsubsection {INFN Legnaro}
A horizontal single-ion microbeam facility for a single-cell irradiations has been designed at the INFN-LNL 7 MV CN Van de Graaf accelerator \cite {report6}. They can deliver light ion beams ($ ^{1}H^{+}$,$ ^{2}H^{+}$,$ ^{3}He^{++}$,$ ^{4}He^{++}$) covering a LET range from 7 to 150keV/$\mu$m and also $ ^{15}N^{++}$. Maximum ion energy achieved is 7 MeV for single charged, 14 MeV for $ ^{4}He^{++}$ and 8 MeV for $ ^{15}N^{++}$. The beam is collimated by a 2-3 $\mu$m diameter using pinholes. Beam is extracted in air through a bi-aluminized mylar window and then collimated at micrometer scale. At Legnaro the ion beam is micro-collimated by means of a tantalum pinhole of 200 $\mu$m thickness with a central hole of 2-3 $\mu$m or 5 $\mu$m.
Cells are irradiated in a stainless steel Petri dish with a 20 $\mu$m thick cell chamber in order to keep them in sterile and humid conditions. The bottom and the cover of the cell chamber. Cells are grown attached to a mylar foil base (7 $\mu$m thick).
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.40]{gfx_infn/setup2}
  \caption{Set up of the horizontal single-ion beam facility at INFN Legnaro.\label{fig:setup2}}
\end{figure}
A high resolution and high sensitivity custom made cooled-CCD camera (Pisa) is used as a particle detector to monitor the spatial particle distribution in air. The window before the pixel matrix is made of a 3 $\mu$m havar foil. A Silicon Surface Barrier (SSB) detector is placed downstream the Petri dish to check the number and the energy of particle crossing the sample. It is used also to check the energy spectrum, the collimation and the beam quality of the beam before irradiation. It is coupled to a CCD camera that allows the beam monitoring \cite {report25}.
When the pre-set number of particle is delivered, an electrostatic deflector switch the beam off. It is placed 5 m before the mylar extraction window and it has a response time of 150 ns.
A post-irradiation off-line check of every particle impact points and individual hit cells is made with a solid state nuclear track detector, placed before the cell sample during the single-ion single-cell irradiation. The control system for a sample movement, single-cell visualization and micropositioning is done by using a inverted in-line phase contrast microscope, a CCD camera and a software program (CELLView). They are able to deliver 1 particle per second.
They also have an other accelerator with which they deliver experiments in radiobiology: the Tandem-XTU that is a 15 MV Van de Graaff accelerator. The ions available are from 30 MeV per unit mass $ ^{1}H$ to 1.5 MeV per unit mass $ ^{197}Au$.
\subsubsection {SNAKE}
SNAKE facility consists of a 14 MV tandem accelerator that can deliver all kind of ions from proton beside the heavy noble gases \cite {report11}. It is designed to focus protons with energies up to 30 MeV as well as heavy ions up to 200 MeV x $q^2$/A to submicron beam spot. It consists of two stage system where a conventional quadrupole doublet is used as first stage with small demagnification factor and a newly developed superconducting quadrupole doublet produces the main demagnification factor at a second stage 30 m downstream from the object slits.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_snake/snake1}
  \caption{Set-up of the microbeam equipment at the Munich tandem accelerator. \label{fig:snake1}}
\end{figure}
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.50]{gfx_snake/snake2}
  \caption{Technical data of the microlens.\label{fig:snake2}}
\end{figure}
The scan of the beam can be done with a magnetic or an electric scanning unit situated in front of the final focusing lens.
The sample is mounted on a 4-axis goniometer with a step size of 1 $\mu$m. 
They tested a proton beam of 20 MeV and a $ ^{32}S$ beam of 90 MeV and they resolution at full aperture is limited by field gradient up to 350 mT/cm \cite {report10}. 
They delivered experiments with this set up on cell culture. The beam size achieved is of about 16 x 8 $\mu m^{2}$. They are seeded between two 6 $\mu$m thick Mylar foils in a saturated atmosphere \cite {report12} in order to prevent drying of the cells and positioned in close contact to the exit window. It is extracted from the beamline by a 7.5 $\mu$m thick Kapton foil which is mounted on a 1 mm bore nozzle. This configuration allows cell survival for more than 30 minutes.
They produce also single ions by using a fast electrostatic beam deflector about 5 m in front of the divergence slits.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.40]{gfx_snake/snake3}
  \caption{Focal plane of SNAKE for living cell irradiation. \label{fig:snake3}}
\end{figure}
\subsubsection{The SIHF at GSI}
This facility is an adaptation of the Unilac accelerator to radiobiological studies. They work with a horizontal beam. They can deliver all type of ions with energy LET from 1.4 to 11.4 MeV/$\mu$m \cite {report23}.
In this system they combine the function of the vacuum window with a high efficiency detection system. For example a 500 nm and 3 x 3 $mm^2$ $Si_{3}$$N_{4}$ is covered from vacuum by 20 nm of gold and 130 nm of CsI to increase secondary electron production.
They also experiment a 2 mm diameter round and 200 nm thick $Si_{3}$$N_{4}$ window showing more stability than square one.
They delivered some tests with carbon ions (7.3 x $10^7$) of 3.6 Mev/u on the first window and with Fe ions (6.2 x $10^6$) of 4.9 MeV/u showing good stability.
They decide to use 200 nm thick and 2-3 mm diameter $Si_{3}$$N_{4}$ window with about 20 nm of Au layer plus about 30 nm of CsI because the detection efficiency achieved is good both for light and heavy ions and an aiming accurancy of 1 $\mu$m.
As previous said a horizontal beam requires a vertical cell dish and they use a Petri dish (10 mm diameter and 2 mm high). This is a wet-cell system with a thin polypropylene foil on which the cells are grown in culture medium. Covering the dish with a thin glass, they can keep the cells in vertical position.
After an automatic recognition, single ions are targeted onto the bars of the grid. All well targeted particles have to traverse the copper and reach the detector with full energy reduced energy while misplaced ions reach the detector with full energy. 
They can deliver 20 ions in a $\mu$m diameter.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_gsi/gsi1}
  \caption{Targeting system using at GSI. \label{fig:gsi1}}
\end{figure}

\subsubsection{LIPSION}
The facility in Leipzig consists of a 3.5 MV Singletron accelerator that can deliver proton and $\alpha$ with energy LET of about 10-20 keV/$\mu$m \cite {report22}. The beam can be focused by a quadrupolar magnetic lens system down to a 50 nm of diameter for low-current mode. A magnetic scanning system moves the focused beam across the sample according to the desired number, width and velocity of the position steps.
The irradiation platform consists of a tunnel with a rectangular cross section which can be attached on this target chamber. The ion beam is equipped with an ion exit window made of $Si_{3}$$N_{4}$ that is 1 mm x 1 mm in area and 100 nm thick. The particle detector p-i-n diode is placed downstream after the cells. They irradiate cells with protons of 2.25 MeV with a flux of 100 protons. The dose deliver to the cells is about 0.5 Gy.
They delivered other experiments also with $\alpha$. They irradiated cells with 5, 10, 20 or 40 2.2 MeV protons/pixel or with one or two 2 MeV $\alpha$ particles/pixel with a particle rate of about 1000 cps. The dose for protons, calculated to the cell nucleus,was 0.7-12 Gy/nucleus. 
The procedure without the medium took less than 2 minutes.
\subsection {Vertical line}
\subsubsection {PTB}
The PTB's accelerator facility comprises a 3.75 MV Van de Graaff accelerator and an energy variable cyclotron with maximum beam energies of up to 20 MeV for protons and 28 MeV for $ ^{4}He$ ions \cite {report13}. This facility utilises the focusing properties of quadrupole magnets to achieve a beam diameter of less than 2 $\mu$m. The set up consists of a horizontal section of 6.8 m and a vertical section of 5.0 m in length. The beam current can be reduced to a few particles per second using the beam defining slits in the horizontal section.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.40]{gfx_ptb/ptb1}
  \caption{Scheme of the beamline and its components. 1.Incoming beam. 2.Object aperture. 3.Divergence slits. 4.Beam deflection plates. 5. Quadrupole doublet. 6.Slit. 7.Bending magnet. 8.Slit. 9.Anti-scattering-slit. 10.Quadrupole doublet. 11.Vacuum window. 12.Experimental area. 13.Basement floor. 14.First floor.\label{fig:ptb1}}
\end{figure}
The microfocusing lens consist of a doublet quadrupole magnets. A borehole (0.25-1.0 mm in diameter) covered with a polymer foil (Mylar), 5 $\mu$m thick,serves as exit window from the beamline (pressure of <10$^{-4}$) at the atmosphere.
The number of particles per cell is controlled by electrostatic deflection plates which are coupled to the detection electronics and which switch the beam on and off within a time of 10 $\mu$s. The use of a bending magnet in the middle of the PTB beamline causes dispersion in the radial direction. The minimum focus achieved is 1 $\mu$m. The size and the position of ion beam are imaging by using a YAG(Ce) scintillator or a thin scintillating foil (BC 400). Single particle detection during irradiation is achieved by detecting the light emitted the scintillating foil in a photomultiplier (PMT). With a foil thickness between 10 and 50 $\mu$m, all particles of any energy available at the facility are detected with an efficiency almost of 100\%. For higher energies the particles traverse the cell disch and can be easily detected by a silicon detector, abolishing the use of scintillator.
The beam characteristic are checked with a CR39 detector. The beamline is vertical so a specific design of the cell dish and a special procedure for cell preparation is required. The entrance window of the cell dish is 1.5-2.5 $\mu$m thick. The cell dish consists of a flat ring (3.5 mm thick) made of medical steel. The front side of the broad wall of the flat ring is subdivided by a groove of 1.2 mm in width and of 0.25 mm in depth. The surface of outer ring of 3.6 mm in width is covered by a thin film of glue.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.30]{gfx_ptb/ptb3}
  \caption{Cell dish.\label{fig:ptb3}}
\end{figure}
In their first experiments the cells were covered with a foil in Teflon of 25 $\mu$m in thickness. They have irradiated 1000 cells over an area of 2-4 mm$^2$. The cells were attached to the foil in a huminidity chamber to avoid dessication. The irradiation was made with 50, 70 and 100 protons of an energy of 1.86 MeV.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_ptb/ptb2}
  \caption{Performance of the PTB-single ion microbeam.\label{fig:ptb2}}
\end{figure}

\subsubsection {Japan Atomic Energy Research Institute (TIARA)}
A charged particle microbeam facility of TIARA has been optimized to deliver high energy microbeam at the AVF-cyclotron with energy up to 10 Mev/u. They developed a single ion hit technique which was constructed on the vertical beam line of the cyclotron. The beams were collimated to a 10 $\mu$m \cite {report23}. The fundamental technology of single ion hitting was developed at the heavy microbeam system connected to a 3MV tandem accelerator. It was built at the Micro Analytical Research Centre (MARC) in Australia \cite {report21}. This facility consists of a focussed external beam with submicron diameter at the cell, an efficient single ion detection system (secondary emission from the exit window), a fast beam switching, a target holder adapted for the irradiation of cells in vitro, a computer controlled system for cell recognition and positioning.
At Jaeri they can accelerate from C to Fe ions. In particular they delivered experiments with microbeam on cells seeded on a 35 mm diameter dish with a 8.5 $\mu$m thick replaceble Kapton foil with precise numbers of 13 MeV/$\mu$m 20 Ne and 11.5 MeV/$\mu$m 40 Ar with calculated LET values of about 380 keV$\mu$m and 1260 keV$\mu$m, respectively. The collimated heavy-ion beams are extracted into air by a tantalum disk perforated by spark erosion.
They use a 5 $\mu$m microaperture on a 100 $\mu$m disk for irradiating with a precise number of 11.5 MeV/nucleon 40 Ar and 13.0 MeV/nucleon $ ^{20}Ne$ ions and a 20 $\mu$m microaperture on a 200 $\mu$m thick disk for 17.5 MeV/nucleon 20 Ne and 18.3 MeV/nucleon $ ^{12}C$.
The Kapton allows to keep cells hydratated during irradiation (less than 30 minutes).
Particles are detected by a plastic scintillator an a PMT installed on an inverted microscope. They improved also the sample positioning using an automated stage. It consists of an on-line and an off-line monitoring system. The off-line system recognizes the cell samples to be irradiated, where a high precision three axis stage, an optical microscope with a CCD camera, a computer network and a software package are located. The system obtain positional data from the CCD camera and the information are transferred to the on-line system. This is combined with the single ion hit system and can irradiate samples by using a software package (IRRADIATE) according to the sample coordinates obtained by the off-line system. They have tested the beam with a CR39 nuclear track detector taking the whole beam profile.

\subsubsection {National Institute of Radiological Sciences (SPICE)}
A recent development of National Institute of Radiological Science is a single particle irradiation system to cell (SPICE). They construct a microbeam facility in the beamline of their Tandem accelerator (3.4 MeV for protons and 5.1 MeV for $\alpha$)\cite {report15}. The focusing of the beam is achieved by the use of a triple Q magnet that can reach a beam spot of less than 2 $\mu$m. The irradiator consists of a dish holder positioning system, a beam deflector and a microscope system. The control of the number of particles is done by a PC at the signal from the ion detector that is a scintillator foil. They expect to reach a resolution of 2 $\mu$m and to irradiate 2000 cells/hour.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.50]{gfx_spice/spice1}
  \caption{A schematic diagram of SPICE facility. \label{fig:spice1}}
\end{figure}
\subsubsection {Gray Cancer Institute}
After the experiments with broad beam they are developing a new facility. They can deliver ion beams from protons to helium ions and the energy reached are about 5.7 MeV $ ^3He$. They performed experiments with a $ ^3He^{2+}$ microbeam from a 4MV Van de Graaff accelerator collimated by a borosilicate glass capillary to micron size \cite {report18}. This is 1 mm long with 1,1.5 or 5 $\mu$m bore collimator and it is closed at the exit by a 3 $\mu$m Mylar vacuum window, then a 4 $\mu$m aluminium foil and then an 18 $\mu$m thick scintillator (Bicron BC400) through which the particles must pass before hitting the target.
The cells' positioning is achieved by the use of a CCD camera, an epi-fluorescence microscope and a micro-positioning stage. Cells to be irradiated are attached to the bottom of a dish which have a 3 $\mu$m thick Mylar. The detector replaced the microscope objective and it is placed above the cell dish. They can irradiate 120 cells per minute.
The irradiation is automatically delivered by adjusting the exact number of particles passing through the collimator and each time the cells are moved under the collimator.
A CR39 detector is used to study the positioning and uniformity of the beam.
According to the collimator performance, the target finding and the stage positioning accurancy, the targeting accurancy reached is below 2 $\mu$m.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_gci/gci3}
  \caption{The schematic of Gray microbeam system,including the beam line, image system and computer control system. \label{fig:gci3}}
\end{figure} 

\subsubsection {Surrey Ion Beam Center}
The Gray Cancer Institute and the Surrey Ion Beam Center project consists of building a new facility that can deliver vertical nanobeam \cite {report16}. They want to develop a horizontal line nanobeam system to act as a test bed for the developement of the vertical beam. They will build and use a vertical focused scanning ion nanobeam using a range of ion from hydrogen to argon. They will develop it on the 2 MV tandetron accelerator they had in the Centre. 
This facility will be focused and steered using electromagnetic fields to target areas less than 10 nm spot with single ions. It will be also used in full current mode for ion beam analysis.
Rapid targeted single-ion positioning allows to increase the number of cells irradiated per second. They expect to irradiate 100000 cells per hour.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.50]{gfx_sibc/sibc1}
  \caption{Design of a vertical beam line.\label{fig:sibc1}}
\end{figure}

\subsubsection {RARAF}
At the Columbia University Radiological Research Accelerator Facility a single-ion beam facility has been constructed. They deliver a defined number of particles produced by a 4.2 MV Van de Graaff accelerator covering a LET range from 30 to 220 keV/$\mu$m from hydrogen to iron. They can irradiate about 10000 cells/h \cite {report17}. 
The beam is focused by an electrostatic lens. An integrated computer control program locates the cells, attached in a monolayer to the thin polypropylene base of a cell culture dish, and positions them for irradiation. 
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.45]{gfx_raraf/raraf1}
  \caption{Overall layout of the microbeam facility and the integrated control system.\label{fig:raraf1}}
\end{figure} 
They are trying to improve the beam resolution from 3.5 $\mu$m to 0.3 $\mu$m to allow the use of more densely-ionizing radiations (heavier ion beam). They use a double triplet system consisting of two electrostatic quadrupole triplets. The beam spot area reach is about 2 $\mu$m diameter. The area of the beam can be focused also with collimator. It is composed by a 2 mm collimator located 1.3 m above the final collimation system. The final collimation system consists of a pair of laser-drilled in 12.5 $\mu$m-thick stainless steel foils and separated by a 300 $\mu$m spacer. The limiting aperture of the first foil is about 5 $\mu$m diameter hole. The second aperture which is a 6 $\mu$m diameter hole, acts as an anti-scattering element.
\begin{figure}[!h]
  \centering
  \includegraphics[scale=.50]{gfx_raraf/raraf2}
  \caption{Collimator geometry. \label{fig:raraf2}}
\end{figure} 
During irradiation, an objective is used to image the cells again in overlapping frames to improve the resolution of the positions of the cell nuclei. The XY motion of the microscope stage is controlled by the computer. The computer positions a cell nucleus over the beam aperture and the specified number of particles is admitted through the beam shutter. The charged particles are detected after passing through the cell using a gas-filled ionization chamber mounted on the 40x objective so that the cells can be observed continuously. The beam shutter is a fast electrostatic deflection system allowing the irradiation of each cell nucleus to be quickly terminated after the specified number of particles have been detected. The beam shutter is opened untill the required number of alpha particles are detected (with a solid state detector located above the cell) to have passed though the nucleus. The shutter is then closed, and the next cell moved under the beam.
They can deliver very low dose (0.1 Gy).
\newpage
\addcontentsline{toc}{section}{References}
\begin{thebibliography}{30}
\bibitem{report1}
M. ~Folkard,K.M. ~Prise, B. ~Vojnovic, H.C. ~Newman, M.J. ~Roper, B.D. ~Michael.
\newblock Inactivation of V79 cells by low-energy protons, deuterons and helium-3 ions
\newblock {\em Int.J.Radiat.Biol.}, 69(6), 1996.

\bibitem{report2}
R.J. ~Munson,D.A. ~Bance, A. ~Stretch, D.T. ~Goodhead
\newblock Mutation and inactivation of cultured mammalian cells exposed to beams of accelerated heavy ions
\newblock {\em Int.J.Radiat.Biol.}, 36(2), 1979.

\bibitem{report3}
I.~Testard,L. ~Sabatier
\newblock Assessment of DNA damage induced by high-LET ions in human lymphocytes using the comet assay
\newblock {\em Mutation Research}, 448, 2000.


\bibitem{report4}
I. ~Bailly,C. ~Champion, P. ~Massiot, P. ~Savarin, J.L. ~Poncy, S. ~Crespin, G. ~Alloy, V. ~Jacob, E.~Petibon
\newblock Development of an experimental system for biological studies:scintillation and solid-track detectors as dose monitors
\newblock {\em Nucl. Inst. and Meth.}, B 187, 2002.

\bibitem{report5}
P. ~Scampoli, M. ~Casale, M. ~Durante, G. ~Grossi, M. ~Pugliese, G. ~Gialanella
\newblock Low-energy light ion irradiation beam-line for radiobiological studies
\newblock {\em Nucl. Inst. and Meth.},B 174, 2001.

\bibitem{report6}
S. ~Gerardi,G. ~Galeazzi, R. ~Cherubini
\newblock Single-ion microbeam as a tool for low-dose radiation effects investigations
\newblock {\em Journal of Physics}, 41, 2006.

\bibitem{report7}
C. ~Shao, Y. ~Furusawa, Y. ~kobayashi, T. ~Funayama
\newblock Involvement of gap junctional intercellular comunication in the bystander effect induced by broad-beam or microbeam heavy ions
\newblock {\em Nucl. Inst. and Meth.},B 251, 2006.

\bibitem{report8}
T. ~Honma, D. ~Ohsawa, K. ~Noda, T. ~Iwashima, H.Y. ~Hogawa, Y. ~Sano, E. ~Takada, S. ~Yamada
\newblock Design and performance of a non-destructive beam-profile monitor utlizing charge-division method at HIMAC
\newblock {\em Nucl. Inst. and Meth.},A 490, 2002.

\bibitem{report9}
Ph. ~Moretto, C. ~Michelet, A. ~Balana, Ph. ~Barberet, W.J. ~Przybylowicz, J.P. ~Slabbert, V.M. ~Prozesky, C.A. ~Pineda, G.~Brut, G. ~Laurent, F. ~Lhoste
\newblock Development of a single irradiation system at CENBG for applications in radiation biology
\newblock {\em Nucl. Inst. and Meth.}, B 181, 2001.

\bibitem{report10}
G. ~Datzmann, G. ~Dollinger, C. ~Goeden, A. ~Hauptner, H.J. ~Korner, P. ~Reichart, O. ~Schmelmer
\newblock The Munich microprobe SNAKE: first results using 20 MeV protons and 90 MeV sulfur ions
\newblock {\em Nucl. Inst. and Meth.},B 181, 2001.

\bibitem{report11}
G. ~Dollinger, G. ~Datzmann, A. ~Hauptner, R. ~Hertenberger, H.J. ~Korner, P. ~Reichart, B. ~Volckaerts
\newblock The Munich ion microprobe: characteristics and prospect
\newblock {\em Nucl. Inst. and Meth.},B 210, 2003.

\bibitem{report12}
G. ~Dollinger, V. ~Hable, A. ~Hauptner, R. ~Krucken, P. ~Reichart, A.A. ~Friedl, G. ~Drexler, T. ~Cremer, S. ~Dietzel
\newblock Microirradiation of cells with energetic heavy ions
\newblock {\em Nucl. Inst. and Meth.},B 231, 2006.

\bibitem{report13}
K.D. ~Greif, H.J. ~Brede, D. ~Frankenberg, U. ~Giesen
\newblock The PTB single ion microbeam for irradiation of living cells
\newblock {\em Nucl. Inst. and Meth.},B 217, 2004.

\bibitem{report14}
S. ~Peng, M. ~Folkard, S. ~Gilchrist, R.J. ~Locke, Zengliang Yu, B.D. ~Michael
\newblock Measurements of the targeting accurancy of the Gray Laboratory charged-particle microbeam
\newblock {\em Nucl. Inst. and Meth.}, B 179, 2001.

\bibitem{report15}
H. ~Yamaguchi, Y. ~Sato, H. ~Imaseki, N. ~Yasuda, T. ~Hamano, Y. ~Furusawa, M. ~Suzuki, T. ~Ishikawa, T. ~Mori, K. ~Matsumoto, T. ~Konishi, M. ~Yukawa, F. ~Soga
\newblock Single particle irradiation system to cell (SPICE) at NIRS
\newblock {\em Nucl. Inst. and Meth.}, B 210, 2003.

\bibitem{report16}
K.J. ~Kirkby, G.W. ~Grime, R.P. ~Webb, N.F. ~Kirkby, M. ~Folkard,K.M. ~Prise, B. ~Vojnovic
\newblock A scanning focussed vertical ion nanobeamn: a new UK facility for cell irradiation and analysis
\newblock {\em Nucl. Instr: and Meth.}, 2007.

\bibitem{report17}
G. ~Randers-Pehrson, C.R. ~Geard, G. ~Johnson, C.D. ~Elliston, D.J. ~Brenner
\newblock The Columbia University Single-Ion Microbeam
\newblock {\em Radiation Research}, 156, 2001.

\bibitem{report18}
M. ~Folkard, B. ~Vojnovic,S. ~Gilchrist, K.M. ~Prise, B.D. ~Michael
\newblock The design and application of ion microbeams for irradiating living cells and tissues
\newblock {\em Nucl. Instr. and Meth.},B 210, 2003.

\bibitem{report19}
M. ~Folkard, B. ~Vojnovic,S. ~Gilchrist, K.M. ~Prise, B.D. ~Michael, O.V. ~Belyakov, A. ~Ozols, G. ~Schettino
\newblock The impact of microbeams in radiation biology
\newblock {\em Nucl. Instr. and Meth.},B 181, 2001.

\bibitem{report20}
B.E. ~Fischer, M. ~Cholewa, H. ~Noghuchi
\newblock The impact of microbeams in radiation biology
\newblock {\em Nucl. Instr. and Meth.},B 181, 2001.

\bibitem{report21}
A. ~Saint, M. ~Cholewa, G.J.F. ~Legge, T. ~Kamiya
\newblock Design of a single ion hit facility
\newblock {\em Nucl. Instr. and Meth.},B 130, 1997.
\bibitem{report22}
\newblock Proceedings of the 6th International Workshop: Microbeam Probes of Cellular Radiation Response
\newblock {\em Radiation Research.},161, 2004.

\bibitem{report23}
\newblock Proceedings of the 7th International Workshop: Microbeam Probes of Cellular Radiation Response
\newblock {\em Radiation Research.},166, 2006.

\bibitem{report24}
M. Belli, G. ~Galeazzi, R. ~Cherubini, S. ~Mazzuccato, G. ~Moschini, O. ~Sapora, G. ~Simone and M.A. ~Tabocchini
\newblock Proton irradiation facility for radiobiological studies at 7 MV Van de Graaff accelerator
\newblock {\em Nucl. Instr. and Meth}, A256, 1987.


\bibitem{report25}
S. ~Gerardi,G. ~Galeazzi, R. ~Cherubini
\newblock A microcollimated Ion Beam Facility for Investigations of the Effects of Low-Dose Radiation
\newblock {\em Radiation Research}, 164, 2005.


\bibitem{Ganil} Grand Accelerateur des Ions Lourds, Caen (\url{http://www.ganil.fr})

\bibitem{Ganil2} Grand Accelerateur des Ions Lourds, Caen  (\url{http://www.ganil.fr/ciril})

\bibitem{Ganil3} Grand Accelerateur des Ions Lourds, Caen 
\\(\url{http://www.ganil.fr/research/interdisciplinarity/G4.html})

\bibitem{gsi} GSI, Darmstadt (\url{http://www.gsi.de})

\bibitem{INFN} Istituto Nazionale di Fisica Nucleare, Legnaro (\url{http://www.lnl.infn.it})

\bibitem{University of Naples} University of Naples, Naples (\url{http://www.na.infn.it})

\end{thebibliography}

\chapter {Beam characterization and monitor for cells'irradiation}
\section {Introduction}
The irradiation of cells requires a stable, homogeneous and monitored beam. It is necessary to build a set up to achieve this aim. The stability and homogeneity of the beam is studied using a PIPS (Passivated Implanted Planar Silicon) detector monitor the flux. In the first case the check consists in counting the particles in the beam in a period of time; in the second one by scanning the beam both in X and in Y direction. Beam monitoring requires a detection system able to discriminate between different fluxes and whom energy loss and straggling are negligible.
The first experimental set up used was based on the idea of detecting the X -ray emission induced by a charged particle beam on a thin layer of a material. A Kapton foil was coated with Rhodium and ZnS for this purpose. The first was chosen because it has the highest X-ray emission cross section in the X-flash detecting region, the second was chosen for both its X-ray and visible emission. In fact it emits a photon in of 340nm of wavelength (blue line) and so it can be used both with the X-flash and the PMT which detection region is in the visible range.
\section {Detection of X-rays for beam current monitor}
\subsection {Expected number of X-rays}
Before starting irradiation with the beam, a calculation of the expected X-rays emission was made. This was achieved by Gupix program and also by means of a formula. 
The formula used was:
\begin{equation}
	Y=N_{p}\epsilon_{x}\exp^{-\mu_{g}x_{g}}\int_{E_p}^{0}\sigma_{x}\frac{N_{a}}{A}\frac{dE}{S(E)}\exp^{-\mu x}
\end{equation}
where $N_{p}$ is the number of particles incident on the target,$\epsilon_{x}$ is the efficiency of the detector, $\sigma_{x}$ is the cross section for X-rays production that can be calculated by another Gupix routine (Gucsa), S(E) is the particle stopping power in the material considered, $N_{a}$  is the Avogadro number and A is the atomic weight of the material. The two exponentials referred to the attenuation of X-ray in the detector-target length, where there is vacuum and so it can be neglected, and to the attenuation of X-rays in the target, that has to be taken into account.
According to thin target approximation, this formula becomes:
\begin{equation}
	Y=N_{p}\epsilon_{x}\sigma_{x} N_{c}
\end{equation}
where $N_{c}$ is the atoms concentration in the target and it considers the target thickness and its density. 
X-ray yields obtained are for steradian units so the geometry of the set up has to be considered.
The routine used to determine the cross section is based on an approximation model of charged particle-matter interaction.
An ion beam can interact with the electron cloud of an atom (Coulomb interaction) leaving this atom in an excited state or even ionisating it. After ionisation, many processes can follow like emission of a characteristic X-ray according to the electron vacancy. There are different possibilities in X-ray emission (X-ray line) with different cross sections. The ionisation and the X-rays emission cross section follow the relation:
\begin {equation}
  \sigma_{x} = \omega \sigma_{i}
\end {equation}
where $\sigma_{x}$ is the fluorescence yield for X-ray production. This relation gives information about the probability of an X-ray line emission knowing the ionisation cross section, that is to say the probability that a particle beam with certain energy interacts with the atom of which the target is made. The cross sections can be calculated in different ways according to different scattering theories.  A plane wave Born approximation can be use to describe the interaction between the charged particle and matter and to evaluate the cross sections of this interaction. This model doesn't completely match with experimental cross section so another model can be used to calculate them. The ECPSSR model takes into account some effects not considered by the PWBA model as following:
\begin{itemize}
	\item ·	Coulomb deflection of the projectile in the field of the target nucleus 
	\item Binding energy effect due to the presence of the slow projectile near the atomic orbit during the collision
	\item Relativistic corrections to the wave functions for orbital electrons
	\item Energy loss of the projectile especially at slow velocity
\end{itemize}
The program used to determine directly X-ray yields is called Gupix and in particular one of its routine Guyls is used. Guyls  allows the determination of the concentration of an element from the X-ray yield. Modifying the parameters, simulations of X-ray yields produced from a target of a certain element can be achieved. In the input file some parameters are required such as geometry, incident particle and its energy, layer thickness and composition, particle exit energy that can be determined by SRIM, and charged of the incident beam. When declaring the particle, it is necessary to specify the cross section the program has to use. The model, used to calculate it, is the same discussing above.
The program output file contains the energy and the yield/sr of each line emission; according to the efficiency of the detector, the X-ray yield to take into account is L line emission for Rhodium and the sum of K (Sulphur) and L (Zinc). 
The calculation was done for different layer thicknesses and for the average experimental current used during irradiation.
To evaluate correctly the number of X-rays it is necessary to consider the detector efficiency and the solid angle source-detector. In fact the program gives a yield for an ideal detector with 100\% of efficiency and for steradian units. The detector used has an efficiency that depends on the energy of the X-ray entering the detection region. These values can be obtained from the plot of detector efficiency against incident energy. According to this plot for Rhodium the X-ray line detected is the one at 2,7 keV and for ZnS are: one Zn line at 1 keV and the other for S at 2,3 keV. The correspondent efficiency is 95\%,85\% and 10\% respectively. The results obtained for each calculation are compared below:
\begin{table}[h!]
\begin{center}
\begin{tabular}{|c|c|c|}\hline
thickness Å & X-ray yield/s program& X-ray yield/s formula\\\hline
1,00E+03      &1,76E-01           &1,31E-01       \\
5,00E+03      &8,81E-01           &6,56E-01       \\
1,00E+04      &1,76E+00           &1,31E+00       \\
    \hline
    \end{tabular}
    \caption{X rays emission from a Rhodium layer for p of 2 MeV (program and formula)}
    \label{tab:X rays emission from a Rhodium layer for p of 2 MeV (program and formula)}
\end{center}
\end{table}

\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
thickness Å & X-ray yield/s program   & X-ray yield/s formula\\\hline
1,00E+03	&4,23E-01	&2,61E-01\\
5,00E+03	&2,12E+00	&1,30E+00\\
1,00E+04	&4,23E+00	&2,61E+00\\
    \hline
		\end{tabular}
	\caption{X rays emission from a Rhodium layer for $\alpha$ of 5.5 MeV (program and formula)}
	\label{tab:X rays emission from a Rhodium layer for $\alpha$ of 5.5 MeV (program and formula)}
	\end{center}
\end{table}

\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		 thickness Å & X-ray yield/s program   & X-ray yield/s formula\\\hline
1,00E+03	&2,613E-01	&1,983E+00\\
5,00E+03	&1,307E+00	&9,916E+00\\
1,00E+04	&2,613E+00	&1,983E+01\\
    \hline
		\end{tabular}
	\caption{X rays emission from a ZnS layer for p of 2 MeV (program and formula)}
	\label{tab:X rays emission from a ZnS layer for p of 2 MeV (program and formula)}
	\end{center}
\end{table}

\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		thickness Å & X-ray yield/s program   & X-ray yield/s formula\\\hline
      1,00E+03	&7,525E-01	&5,558E+00\\
      5,00E+03	&3,762E+00	&2,877E+01\\
      1,00E+04	&7,525E+00	&5,558E+01\\
    \hline
		\end{tabular}
	\caption{X rays emission from a ZnS layer for $\alpha$ of 5.5 MeV (program and formula)}
	\label{tab:X rays emission from a ZnS layer for $\alpha$ of 5.5 MeV (program and formula)}
	\end{center}
\end{table}
These values are obtained for 10$^5$ particles/s and for the solid angle of the experimental set up 6.16 10$^{-2}$ steradian.
According to these X-ray yields the number of X-rays increases with thickness both with formula and program. The main difference between the yields obtained is due to approximation on X-ray interaction cross section and in calculation; in fact the program gives directly the X-ray yield while the result with the formula is obtained through different approximation steps. Both the formula used and the program do not take into account the attenuation of X-rays in target thickness. Using NIST, a database containing the attenuation coefficients in many materials, the absorption of X-ray in the layer can be calculated, underlining the importance of this parameter in the evaluation of X-ray yields. 
Considering attenuation, the X-ray yield decreased; for L-line of Rhodium, for example, the number is half the number calculated with the program. To overcome this problem, a thick target approximation was used. In the input file it was necessary to specify the energy loss in the target by the charged particle and this value was obtained using a program called SRIM. It simply requires the particle type, the incident energy and the target composition. The output file gives the energy loss per angstrom. 
X-ray yields obtained with this approximation give more accurate values comparing to the one obtained for thin target. In fact the relation between yields and target thickness is not linear but contains information on the attenuation.
The results obtained for the simulation are tabled:
\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		thickness Å & X-ray yield/s thick   & X-ray yield/s thin\\\hline
      1,00E+03	&1,763E-01	&1,76E-01\\
      5,00E+03	&4,974E-01	&8,81E-01\\
      1,00E+04	&8,661E-01	&1,76E+00\\
    \hline 
		\end{tabular}
	\caption{X rays emission from a Rh layer for p of 2 MeV (thick and thin approximation)}
	\label{tab:X rays emission from a Rh layer for p of 2 MeV (thick and thin approximation)}
	\end{center}
\end{table}

\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		thickness Å & X-ray yield/s thick   & X-ray yield/s thin\\\hline
      1,00E+03	&3,80E-01	&4,23E-01\\
      5,00E+03	&2,75E+00	&2,12E+00\\
      1,00E+04	&6,32E+00	&4,23E+00\\
    \hline
		\end{tabular}
	\caption{X rays emission from a Rh layer for $\alpha$ of 5.5 MeV (thick and thin approximation)}
	\label{tab:X rays emission from a Rh layer for $\alpha$ of 5.5 MeV (thick and thin approximation)}
	\end{center}
\end{table}

\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		thickness Å & X-ray yield/s thick   & X-ray yield/s thin\\\hline
      1,00E+03	&2,36E-01	  &2,613E-01\\
      5,00E+03	&9,052E+00	&1,307E+00\\
      1,00E+04	&1,556E+01	&2,613E+00\\
    \hline 
		\end{tabular}
	\caption{X rays emission from a ZnS layer for p of 2 MeV (thick and thin approximation)}
	\label{tab:X rays emission from a ZnS layer for p of 2 MeV (thick and thin approximation)}
	\end{center}
\end{table}

\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		thickness Å & X-ray yield/s thick   & X-ray yield/s thin\\\hline
      1,00E+03	&6,61E-01	  &7,525E-01\\
      5,00E+03	&1,637E+01	&3,762E+00\\
      1,00E+04	&2,792E+01	&7,525E+00\\
    \hline
		\end{tabular}
	\caption{X rays emission from a ZnS layer for $\alpha$ of 5.5 MeV (thick and thin approximation)}
	\label{tab:X rays emission from a ZnS layer for $\alpha$ of 5.5 MeV (thick and thin approximation)}
	\end{center}
\end{table}
Results obtained for thin and thick target show a different trend for Rhodium and for ZnS. In fact it seems that yields for thin target are higher than that for thick for Rhodium and it's the contrary for ZnS. This is probably due to the fact that the thick approximation requires the declaration of a parameter that for ZnS over-estimate the real value. Comparing the yields obtained from thick target with the one obtained from formula their agreement is higher than the one shows for thick and thin target. 
According to X-ray theory and to Gupix calculation, X-rays are emitted in each direction with the same yield and they can interact with the culture under irradiation delivering unwanted dose.
Using Nist database the attenuation of X-ray in Kapton and in water can be determined. In fact an X ray produced in the layer has to go through an 8 $\mu$m Kapton foil and then it enters the culture medium where there are the cells. 
Emissions under exam in this study are the 1 keV line of Zn, the 2.3 keV line of S and the 2.7 keV line of Rhodium. After the Kapton foil, for example, just 10\% of X rays emitted from S passed through the layer and 20\% of these are stopped in the culture medium with cells. In these calculations cells were treated as thin layer of 5 $\mu$m thickness of water. The dose delivered is determined by this formula:
\begin {equation}
\dot{D}=\frac{\mu}{\rho}\cdot \varphi \cdot E
\end {equation}
where $\varphi$ is the flux of X ray per seconds and cm$^2$, determined with the thick target approximation of Gupix according to experimental geometry and considering the attenuation, and $\mu$ \ $\rho$, the absorption coefficient, that contains the information of the probability that a photon of a certain energy is absorbed in a layer of material. The attenuation is given by:
\begin {equation}
I(x) = I_{0}\exp^{-\mu x}
\end {equation}
At higher energies X-rays go through the kapton layer without interacting with it. For Sulphur and Rhodium lines the attenuation in Kapton is 30\% and 10\% respectively and 6\% and 10\% of these stops in water. 
According to dose calculation the energy released at cells level by X-ray is lower than $\mu$Gy and so the damage induced is negligible. 
Lines considered up to now are the one that can be detected in experimental condition. In fact the detector used has lower efficiency increasing with the energy of the photon and so they are not useful for monitoring the beam. However they have to be taken into account for an accurate estimation of the potentially damage induced by these X-ray emissions. Rhodium emits X-rays of 20 keV with a low yield (determined with Gupix). From the calculation of attenuation in Kapton and in water made for this energy the percentages of X-rays stopping in them are 0.05 and 0.04 respectively. Dose delivered are too small to induce a considerable damage.
\subsection{Experiment}
After the deposition of a thin layer on a Kapton foil, we have checked if it was possible to use it as a beam current monitor during cell irradiation. The set up consists of an X-flash detector and a PIPS detector. It was built adapting an ERD (elastic recoil diffusion) chamber, used for depth concentration profile determination, to this study. The detector used is an energy dispersive X-ray detector based on the silicon drift chamber principle (SDD).
\begin{figure}[htbp]
\begin{center}
      \includegraphics[scale=0.9]{gfx/sdd.png}
      \caption{Cross section of a SDD detector chip}
      \label{fig:gfx/sdd}
\end{center}
\end{figure}
The detector crystal is moderately cooled by vibration free thermo-electric coolers; any generated heat is dissipated by unforced convection without the need for external cooling means. A super thin radiation entrance window separates the sensitive detector area from the ambient atmosphere and guarantees a good transmission for the X-rays of interest. The main advantage is that it counts X rays fast and it is stable from 1 count to 100000 counts per second.
\begin{figure}[htbp]
\begin{center}
      \includegraphics[scale=0.9]{gfx/sdd2.png}
      \caption{X-Flash working principle}
      \label{fig:gfx/sdd2}
\end{center}
\end{figure}
The geometry of this set up is built to detect the X-ray emitted from the layer and to count the number of particles of the beam as shown in the picture below. The X flash is positioned in front of the layer forming an angle of 10° with the normal of it (beam direction) and an elastic recoil detector placed behind the layer to determine the particle beam flux. The X-flash was calibrated using a $ ^{27}$Co positioned in place of a kapton layer coated with Rhodium or ZnS. X-Flash information is sent to an Analogue to Digital Converter. The signal information is then stored in a memory location of a multichannel analyser and the number of counts vs channel is displayed. Analysing the spectrum obtained, the $ ^{27}$Co peaks can be recognised and also many other peaks due to the backscattering from the detector, to X-ray diseccitation of a $ ^{26}$Fe created by electron capture of a $ ^{27}$Co nucleus are seen.
\begin{equation*}
$ ^{27}$Co + e$^{-}$\rightarrow $ ^{26}$Fe* \rightarrow $ ^{26}$Fe + $\gamma$
\end{equation*}
The energy calibration obtained for the X-flash detector was:
\begin {equation*}
 Energy = 0.0175*Channel + 0.0953
\end{equation*}
\begin{figure}[h!]  
\begin{center}
     \includegraphics[scale=0.1]{gfx/set up.jpg} 
     \caption{Experimental set up}\label{fig:gfx/set up}
\end{center} 
\end{figure}
Layers were irradiated with a proton beam of 2 MeV and an $\alpha$ beam of 5.5 MeV in order to understand if it was possible to use them as beam monitor profile. According to simulation results, the solid angle of the experimental set up was increased (13.9 msr)in order to detect more X-rays.
The results obtained for the irradiation with a 2 MeV proton beam of a Kapton foil coated with a thin layer of Rh and a layer of ZnS are tabled:
\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|c|c|}\hline
		Element& N particles/s& Beam current& thickness & X-ray/s detected\\
		        &             & $\mu$C/s    & Å\\\hline
     Rh     &7.92E+05	    &1.27E-07	    &1.00E+04   &3.92E-01\\
     Rh     &7.45 E+05	  &1.19E-07	    &5.00E+04   &7.73E-01\\
     ZnS    &3.86 E+05	  &6.18E-08	    &5.00E+03   &1.4722E-01\\
		\hline
		\end{tabular}
	\caption{X rays detected for p of 2 MeV}
	\label{tab:X rays detected for p of 2 MeV}
	\end{center}
\end{table}
The same layers were irradiated also with a 5.5 MeV $\alpha$ beam obtaining the following X-ray yields:
\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|c|c|}\hline
		Element& N particles/s& Beam current& thickness& X-ray/s detected\\ 
		       &              & $\mu$C/s    & Å\\\hline
    Rh     &2.62 E+05     &4.19E-08     &1.00E+04  &7.57E-01\\
    Rh     &3.86 E+05     &6.18E-08     &5.00E+03  &8.59E-01\\
    ZnS    &5.70 E+05     &9.12E-08     &5.00E+03  &5.61E-01\\
    \hline 
		\end{tabular}
	\caption{X rays detected for $\alpha$ of 5.5 MeV}
	\label{tab:X rays detected for $\alpha$ of 5.5 MeV}
	\end{center}
\end{table}
The flux of particles corresponds to a dose in water that can be determined using the linear energy transfer according to this formula:
\begin{equation}
\dot{D}=\frac{k}{\rho}\cdot \varphi \cdot LET
\end {equation}
where $\varphi$ is the flux of incident particles, $\rho$ is the density of the material considered and LET is the linear energy transfer of the charged particle in matter. This value varies with the energy of the incident particle and contains the information about the energy loss per unit length by the particle considered and by its secondary emission. SRIM was used to determine it obtaining (in water) 100 keV/$\mu$m and 140 keV/$\mu$m for proton and $\alpha$ respectively.
The range of doses interesting for our study goes from 0.1 Gy to 10 Gy. From the data obtained can be determined the number of X-rays produced for these dose values.
\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|c|}
		\hline
		Particle &Element   & thickness   & X-rays/s for 0.1 Gy\\		
		         &          & Å           &                    \\\hline
p            &ZnS       &5000         &0.095\\
	           &Rh        &5000	       &0.45\\
             &Rh        &10000	       &0.919\\\hline
$\alpha$     &ZnS       &5000         &0.0625\\
	           &Rh        &5000	       &0.125\\
             &Rh        &10000	       &0.21\\
    \hline
		\end{tabular}
	\caption{X rays emissions for 0.1 Gy (p and $\alpha$)}
	\label{tab:X rays emissions for 0.1 Gy (p and $\alpha$)}
	\end{center}
\end{table}
For 10 Gy these yields are 100 times greater but the beam variation is difficult to evaluate and so this set up is not suitable for beam monitor profile, because it is difficult to discriminate between different fluxes.

\section{Experiments with PMT}
\subsection{Detection of photons emitted by ZnS and Ne102 scintillator}
Another detection system developed consists of 10 $\mu$m NE102 scintillator coupled with a photomultiplier. The PMT is supplied with a 4.5 V voltage. The signal is then sent to an ADC and can be seen by means of a multichannel analyser.
A particle hitting the scintillator induces an excited level in the detector that emits photons in the visible spectrum. They are converted in electrons by the photocathode and then multiplied by the dynodes inside the PMT. This detector is more efficient compared to the X flash.
In the first set up the photocathode area was reduced by means of an optic fibre and an experiment was delivered detecting the scintillator light induced by a particle beam.
A background measure was taken without the beam obtaining 25 counts/s and irradiating with the beam obtaining 32 counts/s. To increase photons detected with this set up, the optic fibre was removed and 2 experiments were made: one with scintillator and the other detecting emission from a ZnS foil with a PMT. According to this set up the detector's area was increased achieving a higher number of counts. In this way also the number of background photons increased so we decided to take a measure of the background before each irradiation. The beam fluxes were determined by means of a PIPS detector and the value obtained was compared to the number of photons produced in the scintillator+PMT from the beam.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/graph1} 
      \caption{Photons produced on a Ne102 scintillator by different p fluxes}
      \label{fig:gfx/graph1}
 \end{center} 
\end{figure}

According to the plot, the first data we have taken give a ratio between protons of the beam and photons produced in the PMT of 0.25 for different particle flux. The measurements for the same fluxes were taken again showing a ratio of 1.28 (the slope of the straight line), so two graphs were obtained: one that we called high background, where the first measurements were plotted and the low background, where all the other measurements were plotted. This is probably due to the high value of background that hides the real number of protons.
In the second set of data, two tendencies can be underlined. We think that this is given by the voltage applied to the PMT that leads to bad polarization. These results show that photons, produced by each proton of the beam, are almost 1. According to this preliminary study, the stability and the monitoring of the beam can be checked by means of scintillator + PMT.
The same measurements were taken changing the scintillator foil with a 10000Å ZnS layer coupled with the PMT. The ZnS emissions are detected by the photo-cathode and acquired by the multichannel analyser.
The experiment was delivered without the optic fibre in order to achieve a higher solid angle and so to detect a greater number of photons. Photons produced in the PMT for each ZnS emission caused by a proton beam, are very low. We calculated that for each proton 0,03 photons are expected. We can conclude that ZnS is not suitable to achieve a good beam current monitor.
\section {Energy loss and straggling}
A good beam current monitor requires energy loss and straggling that can be neglected. These parameters can be calculated theoretically by using TRIM that is a software that simulates the interaction between charged particles and the atoms composing the target. In particular it takes into account the trajectory of each ion. The behaviour of protons and $\alpha$ in matter was simulated for 2 different targets: a plastic scintillator (Vinyl Toluene) and a layer of Rhodium on 8 $\mu$m Kapton foil. The simulation were delivered for 2 MeV proton beam and 5.5 MeV $\alpha$ beam with 1000 ions hitting different target thickness that was increased until the beam stops. The output file contains the information about the energy of each ion after the target and about its position. 
The energy loss for an alpha beam in a plastic scintillator was obtained subtracting mean of all ions energy to initial energy. The straggling was determined as the mean deviation of energy obtained from TRIM. The other simulation was done for different thickness of a Rhodium layer on 8 $\mu$m Kapton foil.
\begin{figure}[htbp]
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph2} 
      \caption{Energy loss for p of 2 MeV for different thickness of  Vinyl Toluene scintillator}
      \label{fig:gfx/Graph2}
 \end{center} 
\end{figure}

\begin{figure}[htbp] 
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph3} 
      \caption{Energy loss for $\alpha$ of 5.5 MeV for different thickness of Vinyl Toluene scintillator}
      \label{fig:gfx/Graph3}
 \end{center} 
\end{figure}
 
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph4} 
      \caption{Energy loss for p of 2 MeV for different thickness of Rh}
      \label{fig:gfx/Graph4}
 \end{center} 
\end{figure}

\begin{figure}[htbp] 
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph5} 
      \caption{Energy loss for $\alpha$ of 5.5 MeV for different thickness of Rh}
      \label{fig:gfx/Graph5}
 \end{center} 
\end{figure}
According to the results of simulation the energy loss increases with the thickness in a linear way until it reaches the maximum energy loss for which TRIM stops the calculation.
The energetic straggling obtained for the scintillating foil increased with the thickness in a quite regular way except for 3 and 10 $\mu$m. For rhodium on kapton the trend is not regular and shows many peaks.
\section {Characterization of a scintillator foil for beam current profile}
After this experiment we decided to use scintillator foil to monitor the proton and alpha beam delivered from the accelerator. The aim of this experiment is to calibrate the detector with a proton beam. The calibration is obtained determining the correspondence between the channel in which is contained the information from the detector and the energy of the incident beam. A multichannel analyzer gives information on the number of counts per channel. Two proton beams of different energies where detected and according to the results obtained, the calibration of the junction was achieved:
\begin{equation*}
Energy (kev) =1.788 (keV/channel)*Channel+13.714 keV
\end{equation*}
From the equation given for the calibration it can be determined the energy of any other proton beam or the energy of a beam that goes through a certain material. Approximating peaks to Gaussian ones, information on the beam energy straggling can been obtained from the FWHM. For the detector the energy straggling was of 5 channels for both peaks that correspond to a energy of 23 keV. If a thin layer is positioned in the beam path just in front of the junction, energetic information can be achieved.
We delivered experiments with a 10 and 20 $\mu$m thick Ne102 scintillator, trying to obtain a value of the energy loss and the straggling of the incident beam. For 10 $\mu$m of Ne102 the energy loss is 78 keV with a straggling of 14 keV, while for 20 $\mu$m is 142 keV and 36 keV respectively. Increasing the thickness also the energy loss and straggling increase maintaining the same ratio of the 2 thicknesses.
\subsection {Proton beam}
Ne102 scintillator was tested to understand if it is suitable for beam monitor profile. Two different thicknesses were tested: 10 $\mu$m and 20 $\mu$m. The scintillator was coupled to a PMT to count the number of photons produced by a proton beam of 2 MeV. The beam flux is checked by means of an elastic recoil detector previously described. Before starting each measurement the background value from the PMT was taken. This value is subtracted from the counts obtained from PMT while irradiating with the beam. This allows obtaining the number of photons induced in the scintillator by the transit of the beam. The number of protons is given from the junction. The results for 10 $\mu$m and 20 $\mu$m thick Ne102 scintillator are plotted.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.3]{gfx/Ne102protoni} 
      \caption{Photons produced by 2 thicknesses of Ne102 for different fluxes and relative dose rate}
      \label{fig:gfx/Ne102protoni}
 \end{center} 
\end{figure}
From this plot important information come from: the coefficients of the line of tendence show the number of photons produced by each photon hitting the scintillator. For 10 $\mu$m this value is of about 0.4 that is to say that for 2 incident protons a photon is produced. For 20 $\mu$m 0.85 is obtained so for each proton is produced about 1 photon.
According to this the 20 $\mu$m thickness is more suitable for monitoring the beam because the beam interacts more. The main problem is the energy loss from the beam in its path. This was calculated in the previous paragraph and it's 7 \% of the incident energy. 
These tests allow us to say that the 20$\mu$m scintillator is suitable for beam monitor profile even if the energy loss is not negligible.
The plot shows also the dose rate as a function of the number of incident particles. This value was calculated according to SRIM calculation of protons going through water (approximation used to treat dose to cells). This plot allows knowing the dose given to cells and the number of photons produced for each flux considered. In this calculation the energy loss in the scintillator has to be taken into account to consider the real energy at cells entrance.
\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}
		\hline
		Thickness &Energy loss &Energy at cells'entrance\\		
		$\mu$m    &keV         &keV\\
\hline
		10        &171,97      &1828,03\\
		20        &356,41      &1643,59\\
\hline
		\end{tabular}
	\caption{SRIM calculation of energy loss from p of 2MeV  in Vinyl Toluene}
	\label{tab:SRIM calculation of energy loss from p of 2MeV  in Vinyl Toluene}
	\end{center}
\end{table}
As previous said, it is interesting the study on the beam flux variation. This allows to extend the range of ions hitting the cells and in this way the dose rate can be increased.
The variation in beam flux is achieved by changing the stripper value and so the percentage of the gas contained in the accelerating chamber. This involves a change in the number of negative ions stripped into positive ones.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph6} 
      \caption{p flux as a function of stripper gas \%}
      \label{fig:gfx/Graph6}
 \end{center}
\end{figure}
\subsubsection {Beam homogeinity check}
One of the main problems for irradiation purpose is beam stability and homogeneity of it. We checked beam flux in both X and Y direction by moving the detector each 1 mm in an area 1cm$^{2}$ and taking the number of particles in each position. The scan was obtained by means of a stepper motor controlled from a Pc (LabVieW code).
The proton beam shows good homogeneity in X direction and a worse one in Y direction. For each profile at X fixed the counts increased. The protocol used is time consuming that is to say that requires 30 min to be done and so the beam properties could change during irradiation. 
The whole 3 D profile obtained for the proton beam was:
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/protoni.png} 
      \caption{ 3D profile of a p beam of 2 MeV (kapton exit window)}
      \label{fig:gfx/protoni.png}
 \end{center}
\end{figure}
\subsection {$\alpha$ beam}
After the measurements taken for proton beam monitor profile, it was tested to monitor $\alpha$ beam. For these experiments we used BC400 as scintillator because its composition is similar to the NE 102. In fact they are made of polyvinyl toluene and organic fluors and their chemical formula is $C_{10}H_{11}$. The calibration of the junction was done with different energies of $\alpha$ beam and it was obtained this relation between channel and energy:
\begin{equation*}
Energy (keV)= 7.2702 (keV/channel)*Channel-75.273 keV
\end{equation*}
The signal from the detector was amplified, sent to a shaper and then to an analogue to digital converter. The signal obtained is then displayed by means of a multichannel analyser that make the information from the ADC corresponds to a memory location (channel).
The result obtained from calibration allows achieving energy information from the detector. Two different thicknesses of BC400 were tested: 10$\mu$m and 20 $\mu$m. Each time it was checked the energy loss from the beam in the scintillator. This was achieved by determining the channel corresponding to the peak maximum from the Gaussian fit of the peak and subtracting this number to the initial energy of the beam.
The results obtained for 10 $\mu$m and 20 $\mu$m are 681 keV and 1478 keV energy loss respectively. 
The same measurements were taken also with the Ne102 scintillator. The energy loss in this type of scintillator (800 keV) is lower to the one in BC400.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Spectra} 
      \caption{ Spectra obtained for Ne102 20$\mu$m and BC400 10 and 20$\mu$m}
      \label{fig:gfx/Spectra}
 \end{center}
\end{figure}\\
From this plot the 3 scintillators can be compared. According to Bicron datasheet Ne102 and BC400 are equivalent and so the 20 $\mu$m should be similar to 20 $\mu$m BC400 not to 10 $\mu$m as shown in the graph. For each scintillating foil background and beam measurements were taken. The scintillator is coupled to a PMT that provides the counts from photons incident on the photocathode induced from the transit of the beam. The $\alpha$ beam is monitored with the detector previously calibrated. In this way it is possible to determine the number of photons corresponding to each $\alpha$ hitting the scintillator.
The aim of this experience was to check the ratio photons/$\alpha$ for different beam fluxes.
The number of particles in the beam can be changed by modifying different parameters in the accelerator set up. In particular the stripper gas percentage was changed. This involves a change in the number of negative ions (sorting from the charge exchange channel) turning into positive by means of the stripper gas.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph7} 
      \caption{ Spectra obtained for different stripper \% on 10 $\mu$m BC400}
      \label{fig:gfx/Graph7}
 \end{center}
\end{figure}

\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph8} 
      \caption{ Spectra obtained for different stripper \% on 20 $\mu$m BC400}
      \label{fig:gfx/Graph8}
\end{center}
\end{figure}
The trends show instability in the beam in fact different stripper values correspond to different flux (counts) so the energy shift (35 keV in 10 $\mu$m and 70 keV in 20 $\mu$m ) can be due to a beam not monochromatic. This involves problems with calibration because the energy values taken have a great spread and so the monochromaticity is not certain and so the channel too.
Another way to increase or decrease the flux is by modifying the lithium temperature in the charge exchange channel. This involves an increasing or a decreasing in lithium and of charge exchange, allowing higher or lower beam particle fluxes.
The trend of particle flux with stripper variation is plotted in blue while the HV3 variation increasing the stripper gas is in green. This pressure is the one measured at the exit of the acceleration chamber and it can affect the flux of particle. At the centre of the Tandetron stripper gas allows changing the charge of particles for the second step of the acceleration. Increasing value of stripper corresponds to increased values of particles created and lower values of vacuum. These trends show good agreement with the expected results.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.3]{gfx/Graph9} 
      \caption{$\alpha$ flux and HV3 as a function of stripper gas \% (BC400 10$\mu$m)}
      \label{fig:gfx/Graph9}
\end{center}
\end{figure}

\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.3]{gfx/Graph10} 
      \caption{$\alpha$ flux and HV3 as a function of stripper gas \% (BC400 20$\mu$m)}
      \label{fig:gfx/Graph10}
\end{center}
\end{figure}
After this another measurement was taken with a NE 102 of 20 $\mu$m to monitor photons induced by the $\alpha$ beam and it was compared to the one previously obtained. 
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.3]{gfx/Graph11}
      \caption{Photons produced by BC400 10 and 20 $\mu$m,Ne102 20$\mu$m  for different fluxes and relative dose rate}.\label{fig:gfx/Graph11}
\end{center}
\end{figure}

This plot shows a linear trend only for the Ne102 while for the others trends are not so regular.
According to the slope obtained from linear fit it is possible to know the number of photons for each proton. In 10 $\mu$m scintillating foil it was 0.84 in the first measurements and 1 in the second one.
In 20 $\mu$m this number was 2.2 and 2.6 in 2 different measurements. As it was expected the number increased linearly with thickness. According to these values for the same stripper percentage (flux) two different counts were obtained showing instability in the beam.
Dose rate (Gy/s) is also plotted as a function of the number of incident particles so that is possible to obtain a direct comparison between the flux, the type of scintillator and its rate of photon production and the dose to the cells. The dose rate is determined according to energy loss, as for protons, in the scintillator and then in water.
\begin{table}[h!]
\begin{center}
		\begin{tabular}{|c|c|c|}\hline
		Thickness &Energy loss &Energy at cells'entrance\\		
		$\mu$m    &keV         &keV\\\hline
		10        &834         &4666\\
		20        &1791        &3709\\
\hline
		\end{tabular}
	\caption{SRIM calculation of energy loss from $\alpha$ of 5,5 MeV in Vinyl Toluene}
	\label{tab:SRIM calculation of energy loss from $\alpha$ of 5,5 MeV in Vinyl Toluene}
	\end{center}
\end{table}
\subsubsection {Beam homogeinity check}
For Ne102 the spectra obtained for the junction and for 4 different stripper percentages (flux) were compared. The peaks superposed for three values of stripper, for the other one it was observed a shift of 52 keV. It seems that the beam loses much more energy in that condition. 
The spectrum obtained corresponds to the beam going through the detector and so it is affected by the resolution of the instrument that can be represented by a Gaussian function with a FWHM of 15 keV. This can be treated as a transfer function that multiply for the real signal gives the spectrum displayed by the MCA. Calculating the deconvolution of the spectrum with a Gaussian representing the detector response the initial signal can be obtained and so the information interesting for irradiation purpose. According to some deconvolution tests done, the spectrum obtained is noisy. These results underline the instability in the beam. The 3D profile obtained for the alpha beam is:
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/alfa} 
      \caption{3D profile of alpha beam of 5.3 MeV}
      \label{fig:gfx/alfa}
\end{center}
\end{figure}\\
According to this plot the beam seems to be unstable and this was probably due to Lithium temperature. A stability check was achieved by measuring the beam particle flux in a fixed point during 15 minutes obtaining:
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph12} 
      \caption{Beam stability check (20 min)}
      \label{fig:gfx/Graph12}
\end{center}
\end{figure}\\
This plot shows an instable beam and so it is useful to understand where this trend comes from. A check on the exchange channel was made to verify the instability.It was measured the time in which the Lithium is on and the one in which is off and simultaneously the particle flux in this period of time.
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph13} 
      \caption{Lithium stability}
      \label{fig:gfx/Graph13}
\end{center}
\end{figure}\\
From the plot a periodicity in the beam flux can be underline corresponding to Lithium on (increasing flux) and off (decreasing flux).
After waiting some time to reach a quite stable beam, a second calibration was done. The channel of the 1750 kV beam was taken and it was compared to the one previously obtained. The peak was centred on the 727.689 channel with a FWHM of 31.731.This result has a FWHM value slightly different from the first one, while the channel differs 10 units from the other calibration. 
Varying the high energy magnet is possible to chose one particular energy value, so by changing this parameter can be checked the monochromaticity of the beam. The aim was to estimate the energy straggling of the beam exiting the source. 
The results obtained show that the peak corresponding to beam energy shifts varying the HE magnet that is to say that for each selected energy a beam of different energy can be detected. 
Applying a voltage of 1000kV the energy of the beam is given from:
\begin {equation}
E = T_{C} + T_{E} + F_{C} * TV * (1 + N)
\end {equation}
\begin{itemize}
	\item T$_{C}$: Cathode Voltage (source 860) (kV)                3kV
	\item T$_{E}$: Source extraction voltage (kV)                           16kV (358) or 30kV (860)
	\item F$_{C}$: Calibration factor                                       1.00556 
	\item TV : Applied voltage (kV)                                         150 kV to 2.000kV
	\item N : Ion charge                                                    1 to 8
\end{itemize}
The energy corresponding to that voltage is 3022 keV while applying 1750 kV the energy is 5294 keV. The results obtained, varying the HE magnet, are presented in \fig 
\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph14} 
      \caption{Peak shift for HE magnet variation (1000 kV)}
      \label{fig:gfx/Graph14}
\end{center}
\end{figure}\\

\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph15} 
      \caption{Peak shift for HE magnet variation (1750 kV)}
      \label{fig:gfx/Graph15}
\end{center}
\end{figure}\\
The plot shows a shift in the channel and so in the energy. So it can be concluded that the beam is not monochromatic and the calibration previously done was not correct because the energy value corresponding to the channel is not sure. A possible way to calibrate the detector using an alpha beam is by taking the maximum energy for each voltage.
Plotting the data obtained from variation of HE magnet for 1000kV and for 1750kV in function of energy, can be found the well-know relation given by Lorentz equation:
\begin {equation}
\centering
F=q v B
\end {equation}
in which, according to the definition of kinetic energy , the velocity is given by:
\begin{equation}
v=\sqrt{\frac{2E}{m}}
\end {equation}
obtaining
\begin{equation}
B\approx\sqrt{E}
\end {equation}

\begin{figure}[htbp]  
\begin{center}
      \includegraphics[scale=0.4]{gfx/Graph16} 
      \caption{HE magnet as a function of energy of the beam}
      \label{fig:gfx/Graph16}
\end{center}
\end{figure}\\
\end{document}