Fund Project:Project supported by the National Natural Science Foundation of China (Grant Nos. 1690044, 11575293) and the Presidential Foundation of China Institute of Atomic Energy (Grant No. 11YZ201815).
Received Date:29 September 2018
Accepted Date:28 January 2019
Available Online:01 March 2019
Published Online:05 March 2019
Abstract:To obtain a large uniform beam field for proton single event effect (SEE) experiments, the double-ring double scattering method (DDSM) is employed for spreading the 100 MeV proton beam provided by the 100 MeV proton cyclotron at China Institute of Atomic Energy. With the Geant 4 simulations, the fundamentals of the DDSM are further explored, the achieved effect of our DDSM scheme design is presented, and the influences of some possible factors in practice on the produced beam field are discussed. We find that the the outer part of the second scatter plays an important role in enlarging the area of the uniform field and improving its uniformity. We also find that the first scatter and the inner part of the second scatter play a decisive role in determining the proton flux of the uniform area. The scattering between the spread proton beam and the accelerator tube behind the second scatter damages the uniformity and leads the energy of the produced beam field to straggle. Therefore, the tube should be made as short as possible. The size of the initial beam spot on the first scatter affects the produced beam field to some extent. The spot should be focused as much as possible in a circle with a radius of 0.5 cm. At a larger distance, a larger uniform field can be produced due to the spreading of the proton beam along the space. The decrease in the incident proton energy causes the flux and uniformity to decrease, and also leads the energy loss to increase and the energy of the produced proton beam field to straggle. Using our DDSM schematic design, the simulations show that an 8-cm-diameter beam field with a uniformity of ±1.89% can be produced at a distance of 2.4 m, thereby meeting the need for an SEE experiment of a device-level sample, and that a 20-cm-radius beam field with a uniformity of ±5.32% can be created at a distance of 5.0 m, thereby meeting the need for an SEE experiment of a system-level sample of comparable size. By taking into consideration the uniformity and energy straggling, our design is basically applicable to the protons in the 70?100 MeV energy region that the accelerator can provide directly. Keywords:dual-ring double scattering/ single event effects/ beam spreading/ uniformity
全文HTML
--> --> -->
2.双环双散射体扩束原理简述[12,13]如图1所示, 双环双散射系统由两个散射体构成: 第一散射体S1为高密度物质(如Pb, Ta); 第二散射体S2为双环结构, 内盘S2-inner为高密度物质, 外环S2-outer为低密度物质(如Al). 内盘和外环的厚度设置保证束流通过二者时能量损失保持一致, 这使得内盘较薄, 外环较厚, 但内盘对束流的散射更强. 束流经过S1后会有一定的扩展, 形成中心强、边缘弱的高斯分布, 其较强的中心部分再经S2-inner的较强散射, 其较弱的边缘部分再经S2-outer的较弱散射, 最终在测量平面形成较均匀分布. 图 1 双环双散射体结构示意图 Figure1. Arrangement of scatterers of the dual-ring double scattering method (DDSM).
利用Geant4程序考察各散射体分别在双环双散射体扩束方法中所起的作用来进一步研究该方法产生均匀束流分布的原理. 图4为质子流强为1 nA时, 利用Geant4模拟的只有S1存在、只有S1和S2-inner存在、S1和S2全存在三种情况下在测量平面产生的质子注量率分布, 其中质子全部由S1中心一点入射. 易知, 仅有S1存在时, 束流成高斯分布; S2-inner使经S1后较强的中心部分束流进一步散射, 其结果是注量率在测量平面上半径$A$(值为7.94 cm)内陡降, 半径$A$外增加且高于内; S2-outer使得本应到达测量平面上半径$A$以外的较强的这一部分质子束流进一步散射, 其中有一部分经其散射回到半径$A$内, 使更大区域内的注量率与仅有S1和S2-inner时产生的中心区域的注量率大致相同, 从而使得束流在更大范围内形成更均匀的分布. 可见, S2-outer除了能够保持质子束流能量的一致性以外, 还极大地增大了均匀束流区域, 改善了束流的均匀性. 显然, 在该方法中均匀区域的注量率主要是由S1和S2-inner决定的. 图 4 各散射体在均匀束流分布形成过程中所起的作用 Figure4. Role of every scatter in producing a large uniform beam field.
24.2.第二散射体后加速器管道的影响 -->
4.2.第二散射体后加速器管道的影响
S1, S2均放置于加速器管道(内部为真空)中, 一般在S2之后仍然有一段管道, 且在其末端有Ti膜用以隔绝真空. 如果这段管道太长, 可能会对质子束流的扩展造成一定制约, 从而对最终形成的束流分布造成影响. 取管道内径为14 cm、外径为15 cm、材质为Al, 取Ti膜厚度为100 ${\text{μ{\rm m}}}$, 对这节管道取不同长度时造成的影响进行模拟. 经S2扩束的质子束流再经一段距离传播后会有一部分打到管道内壁上, 这部分质子可能会经管道的散射后再打到所考察的DUT位置. 一方面, 这会导致中心区域的注量率增大, 边缘区域的注量率降低, 从而使均匀区域减小、均匀性降低, 如图5(a)所示. 当管道过长, 比如长150 cm时, 管道对质子束流起到一定的准直作用, 束流截面与管道孔径相当. 另一方面, 经管道散射后到达DUT位置的质子能量会有所损失, 这会导致辐照器件的质子平均能量减小、能散增大, 管道越长, 这种影响就越大, 如图5(b)所示. 当管道长度从0 (即没有管道及其末端的Ti膜)增大到150 cm时, 到达DUT位置的质子平均能量从90.16 MeV减小到85.64 MeV, 能散从0.69%增大到17.05%. 质子SEE实验的目的在于获得SEE截面的能量依赖关系, 故辐照到器件上的质子束流能散越小越好. 考虑到管道对质子束流这两方面的影响, 在S2后应使用尽量短的管道, 姑且将其按5 cm计. 在后文的Geant4模拟中, 均考虑了这节管道及其末端的Ti膜. 图 5 入射质子流强为1 nA时, 第二散射体之后采用0, 5, 50, 100, 150 cm长的管道时在DUT位置所产生的质子注量率分布(a)以及质子的平均能量和能散(b) Figure5. Flux distributions (a), average energy and energy straggling (b) of the protons in the produced beam fields at the DUT position with the 0, 5, 50, 100 and 150 cm accelerator tubes behind the second scatter for 1 nA incident proton beams.
24.3.初始束斑尺寸的影响 -->
4.3.初始束斑尺寸的影响
实际情况中打在S1上的初始质子束斑应有一定尺寸, 这会对最终在测量平面上形成的束流分布产生一定影响. 假定初始束斑为圆形, 其半径${r_0}$取0, 0.5, 1.0, 1.5 cm时在DUT位置形成的束流分布见图6. 可见, ${r_0}$越小, 在中心区域的注量率越小, 边缘区域的注量率越大. 显然, ${r_0}$越小, 就会有越多的质子经过散射能力较强的S2-inner的散射, 这比${r_0}$增大所起到的扩束作用要大, 故有此结果. 另外, 这也导致了${r_0}$较大(取1.0, 1.5 cm)时在DUT位置半径6 cm范围之内质子注量率从外到内逐渐减小的趋势, 破坏了该区域的均匀性. 因此, 在质子束流调节过程中要注意避免使打在S1上的束斑尺寸过大, 尽量使其聚焦在半径为0.5 cm的圆形区域内. 图 6 1 nA质子束流均匀打在第一散射体半径为0, 0.5, 1.0, 1.5 cm的圆形区域时在DUT位置所形成的束流分布 Figure6. Flux distributions of the produced proton beam fields at the DUT position with 1 nA proton beams irradiating 0, 0.5, 1.0 and 1.5 cm radius spots uniformly on the first scatter.
$U = \pm \frac{{\max F - \min F}}{{\max F + \min F}}, $
其中$F$为该区域内的质子注量率. (3)式对均匀性的定义比文献[7]更为保守. 定义束流利用率$\eta $为打到DUT位置半径为$r$的圆形区域内的质子数与打到S1上的质子数之比. 按此定义, 上述设计方案可产生一个均匀性为 ± 1.89%、束流利用率为28.77%、半径为8 cm的照射野, 见图7. 在半径8 cm之外, 束流均匀性开始急剧变差. 图 7 在DUT位置半径为$r$的圆形区域内产生的质子束流分布的均匀性与束流利用率 Figure7. Uniformity of the produced beam field and efficiency of beam use within a circle with a radius of $r$ at the DUT position.
24.4.照射野形成距离的影响 -->
4.4.照射野形成距离的影响
S1距DUT位置的距离$L$称为照射野形成距离. 在进行质子SEE实验时, 测试样品的安装有可能使得$L$在一定程度上偏离2.4 m. 另外, 将来条件具备可以有更大的空间进行SEE实验. 故需要对该设计方案在不同$L$位置形成的束流分布进行研究. 由图8可知, $L$越大, 在DUT位置中心区域产生的注量率越小, 大致与$L$成平方反比, 这正说明中心均匀区域的注量率主要是由S1和S2-inner决定的. 另外, $L$越大, 所形成的均匀区域越大, 很大程度上这是由质子束流沿空间的扩展所导致的. 经计算, $L$在2.3—2.5 m之间变动时, 中心区域半径为8 cm的照射野的均匀性在 ± 1.72%— ± 2.02%之间, 能够满足器件级SEE实验的需要. $L$为5 m时, 中心区域半径为20 cm的照射野的均匀性达到 ± 5.32%, 从而可满足面积与其相当的系统级样品的SEE实验的需要. 图 8 1 nA质子束流在照射野形成距离$L$分别为2.3, 2.4, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0 m时产生的不同束流分布 Figure8. Flux distributions of the produced beam fields with different irradiation field formation distances of 2.3, 2.4, 2.5, 3.0, 3.5, 4.0, 4.5 and 5.0 m for 1 nA proton beams.