【作者】 李玲；

【导师】 安竹；

【作者基本信息】 四川大学 ， 核技术及应用， 2021， 博士

【摘要】 电子与物质的碰撞会产生一系列复杂的相互作用过程并释放出X光子,其中波长连续变化的X光谱被称为轫致辐射。自从X射线在一个多世纪前被发现以来,轫致辐射研究的重要性日益增加,原因在于轫致辐射过程的准确描述不仅对理解电子与靶原子之间的相互作用机制有理论意义,还在核聚变、材料科学和放射医学等诸多领域均具有实际应用价值。例如,在核聚变研究中,轫致辐射作为氚β衰变诱发X射线谱技术的重要输入参数之一,它的精确测量和计算很大程度地决定着这种氚分析方法的准确性;而可靠的轫致辐射截面是放射治疗及诊断核医学中精确剂量计算的重要条件。20世纪70年代以前,普遍认为轫致辐射是由靶原子库仑力作用下入射电子的速度变化引起的,并称之为静态或常规轫致辐射(Ordinary Bremsstrahlung,OB)。而后,另一种被命名为极化轫致辐射(Polarization Bremsstrahlung,PB)的辐射机制进入人们的视野,其形成过程与常规轫致辐射不同,是指当入射电子使靶原子发生极化,靶原子中电子引起的连续X射线发射。并且,两类轫致辐射在与入射电子能量的关系、角分布等多个方面都存在差异。近几十年中,许多学者致力于轫致辐射的理论研究。针对常规轫致辐射,Tseng和Pratt等人在1971年左右建立了一套基于相对论分波法的计算方法,之后由此建立起的轫致辐射能谱表和形状函数表是迄今为止最可靠的常规轫致辐射理论结果(OB理论),用于模拟粒子输运过程的蒙特卡罗PENELOPE程序就是以之为理论基础的。相对于常规轫致辐射成熟的理论研究而言,极化轫致辐射的理论研究还需要进一步深入,目前尚无一套完备的极化轫致辐射理论可用,尽管出现了Avdonina和Korol等人的stripped atom approximation(SA)理论,将常规轫致辐射和极化轫致辐射两种机制均纳入考虑,但由于缺乏角分布,且不包含两种轫致辐射幅值的干涉效应,该理论一般仅用于辐射谱趋势的预估和极化轫致辐射贡献的估算。极化轫致辐射是否存在于总的轫致辐射谱中是目前的研究热点,用实验方法结合现有理论结果进行比较来验证这一问题是目前常用的研究手段。轫致辐射实验所用靶材可大致分为气体靶和固体靶两类,固体靶又可分为厚靶和薄靶。极化轫致辐射的存在已经于2003年左右在Quarles等人的电子碰撞惰性气体靶实验工作中得到了明确的证实,而针对固体靶,这一问题还存在许多争议。有多篇文章表明,电子碰撞固体厚靶所产生的轫致辐射实验谱与只考虑常规轫致辐射的理论谱相比大体上相符。尽管在低光子能段存在一些差异,但他们认为固体靶轫致辐射中不存在极化轫致辐射的贡献。然而有部分研究结果显示,极化轫致辐射贡献明显存在于利用β衰变源释放出的电子碰撞多种金属厚靶时产生的轫致辐射谱中,且随光子能量的增加而降低,随靶原子序数的增加而增加。因此,固体厚靶轫致辐射过程中是否有极化轫致辐射的贡献仍需更精确的实验工作加以佐证。对电子在厚靶中多重散射过程的计算是相当复杂的,尽管存在PENELOPE这样的蒙特卡罗模拟计算工具,但由近似处理、模型的准确度以及截面参数的准确度等各种因素带来的误差是无法避免的。利用固体薄靶测量轫致辐射截面值可直观反映入射电子与靶物质产生轫致辐射的概率大小,能够与现有的理论结果进行更直接的比较。理想情况下的薄靶轫致辐射是指一个电子与一个原子发生单次碰撞所产生的辐射,它不受多重散射、电子能量损失等的影响。但由于受到制靶技术的限制,自支撑薄靶的制作是极为困难的,且薄靶的厚度和平整度往往会给实验带来较大误差。1958年至1987年,仅有几篇文章报道了关于50-200 keV电子碰撞固体薄靶产生轫致辐射截面的绝对测量工作,且其测量不确定度高达20%。最近,有研究测量了20-100 keV电子碰撞五种原子产生的轫致辐射双微分截面,对于高入射电子能量(例如,>50 keV)和高Z元素(Te、Ta和Au),实验截面值与Pratt等人的常规轫致辐射理论符合得很好,且在谱的高光子能量尾端,二者的符合程度最好,随光子能量降低,差异逐渐增加。对于低Z元素C,所有入射电子能量对应的理论值系统性地低于实验值,平均偏差约为25%;20 keV入射电子碰撞低Z元素Al的数据也呈现类似差异。初步推测实验结果与常规轫致辐射理论截面数据间的差异可能与极化轫致辐射有关系,但并未给出明确结论。值得注意的是,该研究只给出了较高光子能量范围内的实验结果,例如,对于光子发射角90°,入射电子能量约20 keV的条件下,只给出了光子能量大于10 keV的数据。一些文献曾提出极化轫致辐射最有可能出现在低光子能量区域,因此低光子能量段是值得我们格外关注的。总的来说,高精度的薄靶实验和截面数据仍较为稀少,尤其是低入射电子能量和低光子能区的相关数据就更为匮乏。本文中,我们采取厚靶与薄靶相结合的方法,全面地分析固体靶轫致辐射这一物理现象。首先,我们选取了从低Z到高Z元素共11种高纯度的厚碳(C)、铝(Al)、钛(Ti)、铜(Cu)、锆(Zr)、铌(Nb)、银(Ag)、碲(Te)、钨(W)、金(Au)和铅(Pb)靶,利用扫描电镜提供5-25 keV聚焦电子束并竖直向下入射至靶面,采用带有超薄窗的硅漂移探测器等测量设备,设计偏转磁铁以消除背散射电子对探测器和测量结果的干扰,并优化数据处理方法和探测器效率刻度方法,精确测量轫致辐射光谱。通过将厚靶轫致辐射实验谱与只包含常规轫致辐射截面数据而不考虑极化轫致辐射的蒙特卡罗模拟谱进行对比,发现光子能量大于1 keV范围的能谱均符合得很好。另外,我们还采用薄靶方法,在厚靶实验的基础上改进了实验装置,对5-25 keV能量的电子轰击6种薄碳(C)、铝(Al)、钛(Ti)、铜(Cu)、银(Ag)和金(Au)靶生成的轫致辐射谱进行准确的测量,并结合蒙特卡罗计算,对PENELOPE程序作并行化处理,模拟真实的实验几何结构,对电子多重散射及衬底散射等效应进行了修正,最终得到轫致辐射双微分截面测量值。采用纳米级薄靶加薄碳衬底的制靶方式,并利用卢瑟福背散射方法精确测量靶材料厚度。我们给出了低光子能量范围内(低至1-2 keV)的截面值,更有利于对极化轫致辐射贡献的判断。通过将实验截面值分别与Pratt等人的OB理论值和Avdonina等人包含了极化轫致辐射贡献的SA理论值进行比较,发现对于低Z元素(C和Al),测量值与OB理论值和SA理论值都存在一些差异,尤其是在低光子能量区域,但OB理论值与实验值在幅值和形状上的一致性明显优于SA理论值,且两者的符合程度随光子能量增加而逐渐变好;对于中、高Z元素(Ti、Cu、Ag和Au),实验值与OB理论值符合得很好,而与SA理论值存在显著差异。在本文所研究的5-25 keV入射电子能量范围内,厚靶和薄靶的分析结果均没有出现任何表明有极化轫致辐射贡献存在的证据,解决了近年来关于低能电子碰撞固体靶是否会产生极化轫致辐射的争议,同时也证明了Pratt等人的OB理论可以很好地描述低能电子轰击固体靶产生轫致辐射这一物理过程。更多还原

【Abstract】 The collision of electrons with matter produces a complex series of interaction processes and the release of X-rays,of which the X-ray spectrum with continuously varying wavelengths is known as bremsstrahlung.Since the discovery of X-rays over a century ago,the research on bremsstrahlung has become increasingly important due to the fact that an accurate description of bremsstrahlung is not only of theoretical importance for understanding the mechanisms of interaction between electrons and target atoms,but also has practical applications in many fields such as nuclear fusion,materials science and radiology.For instance,in the study of nuclear fusion,bremsstrahlung is one of the key input parameters to the beta-decay induced X-ray spectroscopy technique,and its accurate measurement and calculation largely determines the accuracy of this tritium analysis method;reliable bremsstrahlung cross sections are a prerequisite for accurate dose calculations in radiation therapy and diagnostic nuclear medicine to ensure the effectiveness of treatment.It is generally accepted that bremsstrahlung is caused by the incident electrons under the Coulomb force of the target atom,and is referred to as ordinary or static bremsstrahlung(OB).It was not until the 1970 s that another radiation mechanism called polarization bremsstrahlung(PB)was first proposed,which refers to the continuous X-ray emission caused by electrons in the polarized target atom.The two types of bremsstrahlung differ in a number of aspects such as the formation process,the relationship with the energy of the incident electrons,and the angular distribution.In recent decades,many scholars have devoted to the theoretical study of bremsstrahlung.For ordinary bremsstrahlung,Tseng and Pratt et al.proposed a theoretical approach based on relativistic partial-wave method around 1971,and the resulting tables of bremsstrahlung energy spectra and shape functions established by Pratt and Kissel et al.are the most reliable theoretical results for ordinary bremsstrahlung to date.The Monte Carlo method PENELOPE for simulating the coupled electron-photon transport processes is based on this ordinary bremsstrahlung theory.In contrast to the mature theoretical studies of ordinary bremsstrahlung,the theoretical studies of polarization bremsstrahlung need to be further developed.The stripped atom approximation(SA)theory by Avdonina and Korol et al.has emerged in late twentieth century,in which both ordinary and polarization bremsstrahlung are considered.But it generally used only for the prediction of the trends and the estimation of the polarization bremsstrahlung contribution due to the lack of angular distribution and the interference of the two bremsstrahlung amplitudes.And there is not yet a complete theory available for polarization bremsstrahlung.Whether polarization bremsstrahlung exists in the total bremsstrahlung spectrum is a focus of research at present,and comparison between the experimental results with the theoretical predictions is a commonly method to verify this issue.The targets used in bremsstrahlung experiments consist of two categories: gas targets and solid targets,while the solid targets include thin target and thick target.The existence of polarization bremsstrahlung in the total bremsstrahlung spectra for gas targets was demonstrated in 2003 in the experimental work of Quarles et al.,but the issue for solid targets is still controversial and inconclusive.Several papers have shown that the experimental spectra of bremsstrahlung produced by electron collisions with thick solid targets were generally consistent with the theoretical results considering only ordinary bremsstrahlung.Although there were some differences in the low photon energy regions,they concluded that there is no polarization bremsstrahlung contribution in the solid target bremsstrahlung.However,some of the results showed that the polarization bremsstrahlung contribution was clearly present in the bremsstrahlung spectra produced by the collisions of electrons emitted from a beta decay source with a variety of metal thick targets,and increased with decreasing photon energy and increasing target atomic number.Therefore,the contribution of polarization bremsstrahlung to the total bremsstrahlung spectra for thick solid targets still needs to be validated through more precise experimental work.The calculation of the multiple scattering process of electrons in thick target is quite complex,and despite the availability of simulation tools such as the Monte Carlo PENELOPE code,there are still errors arising from various factors such as the approximation treatment,the uncertainty of the model and the cross section data in the Monte Carlo simulations.The measured bremsstrahlung cross section values using a thin solid target can reflect the probability of bremsstrahlung generated by the incident electron colliding with the target matter,and allow a direct comparison with the existing theoretical results.Ideally,the thin-target bremsstrahlung is the radiation produced by a single collision of an electron with an atom and is almost free from the influence of multiple scattering and the energy loss of electron.However,due to limitations in target-fabrication techniques,the fabrication of self-supporting thin target is extremely difficult,and the thickness and flatness of which often bring large error to the experimental results.From 1958 to 1987,only a few papers reported absolute measurements of bremsstrahlung cross sections produced by 50-200 keV electron colliding with thin solid targets with uncertainties of up to 20%.Recently,bremsstrahlung cross sections for 20-100 keV electrons have been measured,while only the data in higher photon energy region have been given,e.g.,for the photon emission angle of 90° and the incident electron energy of ～20 keV,only data for photon energies greater than 10 keV were available.The experimental data agree well with the OB theory of Pratt et al.for high incident electron energies(e.g.,>50 keV)and high Z elements(Te,Ta and Au),and the best agreement was found at the high photon energy ends of the spectra.For low Z element C,the theoretical values corresponding to all incident electron energies were systematically significantly lower than the experimental data,with an average deviation of about 25%;the data produced by 20 keV electron with low Z element Al show similar differences.It was speculated that the discrepancies between the experimental results and the OB theoretical values may be related to the polarization bremsstrahlung,whereas no clear conclusion was given.Thin-target experiments and cross-section data with high precision are still scarce,especially for low incident electron energies and low photon energies.In addition,some researchers have suggested that polarization bremsstrahlung is most likely to occur in the low photon energy region,so the attention to bremsstrahlung at low photon energies is of great importance.In this paper,we have conducted a comprehensive analysis for the solid target bremsstrahlung with low-energy electron impact by using the thick-target and thintarget method.Firstly,we selected 11 high purity thick targets from low Z to high Z elements C,Al,Ti,Cu,Zr,Nb,Ag,Te,W,Au and Pb.The bremsstrahlung spectra produced by 5-25 keV focused incident electrons that were provided by a scanning electron microscope have been detected using a silicon drift detector with an ultrathin window.In addition,a pair of 800 G magnets were designed to eliminate the interference of backscattered electrons to the detector,and new method for data processing and efficient calibration ensured the accuracy of the experimental data.By comparing the measured spectra of the thick target bremsstrahlung with the calculated results by the PENELOPE code,which contains only ordinary bremsstrahlung crosssections,it was found that the energy spectra in the range of photon energies above 1keV were in good agreement.Besides,we have improved the experimental setup on the basis of the thick-target experiments,and measured the bremsstrahlung absolute double differential corss sections for six elements C,Al,Ti,Cu,Ag and Au with 5-25 keV electrons impact.The C target was a self-supporting film and the other five were two-layer structure targets.The thicknesses of all the target films were a few tens of nanometers,which have also been measured using the Rutherford backscattering spectrometry method.The Monte Carlo code PENELOPE has been parallelized to simulate the real geometries of the expeirments in order to provide a correction for some effects such as electron multiple scattering and the backscattering in the target substrate.The experimental cross-section data were compared with the OB theoretical values of Pratt et al.and the SA theory taking into account polarization bremsstrahlung,respectively.For low Z elements(C and Al),there are some differences between the measured data and the two theories,especially in the low photon energy region.However,compared with the SA theory,the OB theoretical values are obviously closer to the measured data in both amplitude and shape,and the agreement becomes better with the increase of photon energy.For medium and high Z elements,the measured data are in good agreement with the OB theoretical values,while there are significant differences with the SA theory.The results for both thick and thin targets indicate that interaction between lowenergy electrons with solid targets does not produce polarization bremsstrahlung and that the physical process can be well described by the OB theory.The work in this paper has resolved the recent controversies over whether the polarization bremsstrahlung contributes to the X-ray radiation for solid targets with low-energy electron impact.更多还原