Diamond

The exceptional properties of diamond made it an attractive material in various domains of applications from cutting tools to electronic and optoelectronic applications as thermal conductors as well as jewelleries. Until 50’s years diamond from natural stones was studied also for its potentialities in radiation detection due to its high radiation hardness. The problem of using diamond as radiation detector is the diversity of the sample coming from various stones. A drastic selection is necessary in order to find one exceptional sample able to work as a detector. Despite this, some natural diamond samples are using as spectrometer to calibrate physic research installations or sell as ionisation chamber for use in medical physics. One part of the European project MAESTRO focus on this last point, study of diamond as point dosimeter.… but why diamond?

The aim of the external radiotherapy is to destroy the cancerous cells by irradiation with minimum damage on the safe tissues. The instrumentation used to this purpose consists in electron accelerators or cobalt irradiators which are calibrate to give the dose calculated for each specific treatment. One of the principal goals of a treatment is to control the dose received by the patient during the treatment and the beam calibration. Usually air cavity ionisation chamber are dedicated to this measurement due to their little dimension of the detection volume, weak dependence of the response with respect to the energy of the irradiation beam, linearity of the response with the dose, reproducibility of the measurement below 5% of variation per year. The emergence of new irradiation techniques as IMRT implies improvement of measurements techniques. The juxtaposition of several beam of very little size implies minimizing the size of the detector. Furthermore if the device could be made from tissue equivalent material it will simplify the calculation and minimize the perturbation of the detector response. Concerning the off-line dosimetry with thermoluminescent dosimeter (TLD), characteristics as non-toxicity, chemical stability of the sample are very attractive.

From these whole considerations diamond appears clearly as the perfect candidate. The following table summarise the attractive intrinsic diamond properties for applications as ionisation chamber and TLD.

Density	3.51	High radiation hardness
Resistivity	10 ¹⁶ W.m	Low dark
Band gap	5.5 eV	current
Z atomic number	6	Tissue equivalence (Z=7.4), non toxicity

The high density of the material implies high radiation hardness that able long use without deterioration of the ionisation chamber or TL dosimeter. The resistivity and the band gap of the diamond implying low dark currents at room temperature, allows simple use as solid ionisation chamber without necessity of junction fabrication. The chemical characteristics of diamond are very attractive for ionisation chamber as well as TLD. In the first case, it will simplify the calculation, in the second case it will improved the safety of use due to the non-toxicity of the material, compared to the existing material as LiF. Furthermore it could be sterilisable for in-vivo measurements.

The use of dosimeter for medical application and particularly radiotherapy requires several characteristics. Stability, reproducibility of the signal, linearity with dose and independence of the response versus the dose are required both for ionisation chamber and TLD. Despite this, the specifications in diamond properties required for one or the other point dosimeter are completely different. For example, in ionisation chamber pure material is required to minimize loss of charge created by incident radiation. On the opposite impurity are necessary in TLD to create thermoluminescent signal. Both aspects will be studied in the MAESTRO project WP3.1 collaboration work.

As mentioned previously, selected natural diamond are sell to be used in different fields of radiation detection but one of the principal problems for using natural stones is the diversity. Detection properties of diamond strongly vary with the concentration and the type of the impurities present in the material. For natural diamond, a classification as been established with a separation of stones in two types I and II depending on the concentration and nature of impurity incorporated in the crystal. To summarise: “detector quality” for natural stones, concerns type IIa diamond which represent ~1% of natural stones. This explains the preliminary difficulties for finding one detector able to work in radiation field applications, the delay and the high cost of this kind of ionisation chamber.

Synthetic diamond is a good alternative for natural stones and several techniques are used to grow synthetic samples. In the MAESTRO project the samples studied are obtained mainly by High Pressure High Temperature (HPHT) procedure or Microwave Plasma Chemical Vapour Deposited (MWPECVD) process. In HPHT process samples obtained are monocristalline, of very little size of about 3´3 mm, with residual impurities due to the use of catalyst as nickel. In CVD process, samples could be polycrystalline or monocristalline depending on the substrate on which diamond growth occurs. The samples grown on silicon substrate are polycrystalline and the size could be adaptable, incorporation of impurity could be minimized or voluntary added in controlled concentration depending on the application. Samples grown on diamond itself are monocristalline, of little dimension due to the initial size of the substrate. For ionisation chamber application, both polycrystalline and monocristalline synthetic diamond could be used due to the little dimension of the device required. For TL dosimeter polycrystalline CVD diamond is the best alternative due to the possibility of TL dosimeter fabrication coming from the same substrate.

As mentioned previously, we can distinguish two processes of growth, monocristalline and polycrystalline, but the principle of diamond lattice formation is the same in both case and described below.

Diamond growth is based on dissociation of gas species by microwave excitation. Parameters as gas pressure in the chamber, microwave power, substrate temperature and percent of gas (methane versus hydrogen) are of great importance and have to be optimised. Conditions of growth could vary in the range listed below.

Microwave power	Pressure	Temperature	% CH4	Growth rate
1500-4500 W	70-140 mbar	750-950°C	0.5-4	0.1-4 mm/h

Polycristalline diamond growth occurs on different substrate than diamond itself. Two steps are distinct in the process: the first one is nucleation on defects created at the substrate surface; the second one is the growth itself. After growth the substrate could be mechanically removed and the 2 inch self supported diamond cut in pieces of various size adapted to the application. Monocristalline CVD sample grown on HPHT substrate have to be polished to obtain self supported diamond without contribution of the substrate.

Four samples illustrating the four steps of the growth process, from left to right : silicon substrate, nucleation on silicon, growth of diamond, and self-supported diamond sample

Despite all the advantages of diamond, there are existing problems as the presence of residual impurities or defects in synthetic diamonds. The structural defects or impurities could modify the response of ionisation chamber or TL dosimeter.

These defect levels are present in all forms of synthetic diamonds (e.g. HPHT and CVD) affecting the reproducibility of the dosimetric properties.

The energy depth of the level in the diamond gap varies from sample to sample, thus response can vary from sample to sample significantly. Similar difficulties are also observed in the reference PTW ionisation chambers made from natural diamond, which require daily pre-irradiation prior to use (typically. 8 to 10 Gy) for the improvement of the stability properties (priming).The defect levels influencing the response of the detector could be due both to the bulk and the surface sample. Therefore full understanding of material properties accompanied with optimisation of growth procedures and device set-up are needed to optimise the detector response. The aim of studies carried out by various partners of WP3.1 is to identify ways to improve the device performances even though the diamond detectors contain some defect levels and their properties are not ideal. Of these, several technological steps have to be optimised and can be listed as follow:

CEA-LIST
Diamond growth, optimisation of material properties and evaluation of TL an ionisation chamber potentialities

IFJ
Development of TL and ionisation chamber dosimeters and evaluation of potentialities in clinical environment

DFC and ISS
Development of ionisation chamber dosimeter and evaluation of potentialities with medical radiotherapy apparatus

Several investigations have been performed by different partners of WP3.1 on ionisation chamber characterisation and evaluation of potentialities in radiotherapy field.

IFJ works are focused on the fabrication of an active dosimeter dedicated specifically for radiotherapy application. Different holders have been made: a universal holder used for diamond crystal testing and a holder with a heating system in order to heat sample at elevated temperature (300°C). These holder have great advantages: there is no need to glue the sample to have electrical contact and no effect of air ionisation (which would disturb the measurement) is possible due to the design and the packaging of the holder.

Performances of HPHT diamond crystal have been tested using the universal holder. Test in clinical environment with ⁶⁰Co source and X-rays have been performed : linearity with dose rate, stability and reproducibility have been evaluated showing striking diversity of the detection characteristics for individual devices and showing that growth have to be optimised to follow requirements for radiotherapy applications.

Another series of diamond mounted as ionisation chamber have been tested jointly by DFC and ISS. A new generation of detector grade CVD samples from E6 have been investigated and compared to ionisation chamber.

Current response of CVD vs time after priming (10 MV, 200 MU/min). For comparison the IC current is also reported. A scaling factor was applied to the IC signal.

Beta tests and preliminary clinical tests have been performed. Dosimetric performance of the detector have been measured with a photon beam from a LINAC accelerator and time stability, response dynamic, dose rate dependence and energy dependence have been evaluated. Results for the sample tested give evidence of existing drawbacks for IMRT applications showing that improvement are necessary concerning time stability and dose rate dependence in order to reduce their influence on dose assessment.

Studies of diamond material are currently in progress at CEA. Ionisation properties of CVD diamond made at CEA have been investigated and detector behaviour (in term of reproducibility and stability of the response) correlated with the presence of defects in the material. As it was mentioned previously ionisation chamber on CVD diamond presents some instability: pumping phenomenon which is well known during the first step of irradiation followed by an overshoot which occurs on the second step of irradiation.

Two ways are planned to avoid the influence of trap levels: improvement of experimental set-up and/or improvement of crystal quality. CEA propose a design of tissue-equivalent holder in order to performed measurements in radiotherapy environment with the possibility of heating sample. Boron doped diamond is used as heater and optimised intrinsic CVD diamond as detector.

The other way to improve the ionisation chamber properties are currently studied by improvement of diamond quality with respect to the requirement of radiotherapy field. Monocristalline samples as well as polycrystalline are investigated with respect to the growth conditions.

Some partners of WP3.1 (CEA and IFJ) focus a part of their activity on TL dosimeter application. It is well known that CVD diamonds show good potentialities for TL dosimetry particularly in the field of radiotherapy applications. The presence of deep levels observed on several intrinsic diamonds (CVD as well as natural) at high temperature (range 240-280°C) could be exploited for this application, but there exist some drawbacks as moderated thermal fading, important bleaching and short range TL signal linearity.

TL curve of intrinsic CVD diamond sample with illustration of thermal fading and optical bleaching

The main drawback of using non-doped CVD diamond is the lack of control on the impurities responsible for the presence of deep trap levels. The purpose in this field is to control the trapping level in the material with deliberate incorporation of impurities in the diamond during the growth. Several impurities as nickel or phosphorus are supposed to be good candidate to induce deep trap level in the band gap, but the most common impurity is nitrogen known as residual impurity in intrinsic CVD diamond that could be responsible for the existing uncontrolled deep trap level in these samples.

Nitrogen doped diamond have been grown at CEA-LIST with various [N₂H₂] concentration. Dosimetric properties have been investigated in terms of reproducibility, detection limit, sensitivity and TL signal linearity with the dose. Results show clear modification of the TL curve shape that could be interesting to improve the readout procedure facility (no preheating stage required) and the calculation of the absorbed dose in the diamond dosimeter. The sensitivity of the TL signal tested from various nitrogen incorporations seems to indicate that there exists an optimised nitrogen incorporation value, for which the TL signature is maximised. The linearity of TL signal with the dose for nitrogen-doped sample is improved compared to non-doped sample. Low nitrogen incorporation seems to improve both intensity and linearity, whether a too high nitrogen concentration seems to deteriorate significantly the detection limit and the sensitivity of samples. A compromise is required to optimise both sensitivity and signal linearity with the dose.

Current and further work will focus on the range of energy and dose of radiotherapy treatment for evaluation of TL diamond properties. Others impurities will also be incorporated and their effects on TL properties of diamond estimated.

Left : Modification of the TL curve shape for different concentration of nitrogen