SummaryMost of the work carried out in relation to contrast mechanisms and signal formation in an environmental scanning electron microscope has yet to consider the time dependent aspects of image generation at a quantitative level. This paper quantitatively describes gaseous electron-ion recombination (also known as 'signal scavenging') in an environmental scanning electron microscope at a transient level by utilizing the dark shadows/streaks seen in gaseous secondary electron detector images of alumina (Al 2 O 3 ) immediately after a region of enhanced secondary electron emission is encountered by a scanning electron beam. The investigation firstly derives a theoretical model of gaseous electron-ion recombination that takes into consideration transients caused by the time constant of the gaseous secondary electron detector electronics and external circuitry used to generate images. Experimental data of pixel intensity versus time of the streaks are then simulated using the model enabling the relative magnitudes of (i) ionization and recombination rates, (ii) recombination coefficients and (iii) electron drift velocities, as well as absolute values of the total time constant of the gaseous secondary electron detection system and external circuitry, to be determined as a function of microscope operating parameters such as gaseous secondary electron detector bias, sampleelectrode separation, imaging gas pressure, and scan speed. The results revealed, for the first time, the exact dependence that the effects of secondary electron-ion recombination on signal formation has on reduced electric field and time in an environmental scanning electron microscope. Furthermore, the model implicitly demonstrated that signal loss as a consequence of field retardation due to ion space charges, although obviously present, is not the foremost phenomenon causing streaking in images, as previously thought.

Absolute charge calibration of scintillating screens for relativistic electron detection Rev. Sci. Instrum. 81, 033301 (2010); 10.1063/1.3310275Silicon photodiodes for low-voltage electron detection in scanning electron microscopy and electron beam lithography J.This work investigates the generation and detection of gaseous scintillation signals produced in variable pressure scanning electron microscopy through electron-gas molecule excitation reactions. Here a gaseous scintillation detection ͑GSD͒ system is developed to efficiently detect photons produced via excitation reactions in electron cascades. Images acquired using GSD are compared to those obtained using conventional gaseous secondary electron detection ͑GSED͒ and demonstrate that images rich in secondary electron ͑SE͒ contrast can be achieved using the gaseous scintillation signal. A theoretical model, based on existing Townsend theories, is developed. It describes the production and amplification of photon signals generated by cascading SEs, high energy backscattered electrons, and primary beam electrons. Photon amplification ͑the total number of photons produced per sample emissive electron͒ is then investigated and compared to conventional electronic amplification over a wide range of microscope operating parameters, scintillating imaging gases, and photon collection geometries. These studies revealed that argon gas exhibited the largest GSD gain, followed by nitrogen then water vapor, exactly opposite to the trend observed for electron amplification data. It was also found that detected scintillation signals exhibit larger SE signal-to-background levels compared to those of conventional electronic signals detected via GSED. Finally, dragging the electron cascade towards the light pipe assemblage of GSD systems, or electrostatic focusing, dramatically increases the collection efficiency of photons.

The Environmental Scanning Electron Microscope (ESEM) is capable of image generation in a gaseous environment at sample chamber pressures of up to 20 torr. in an ESEM, low energy secondary electrons emitted from a sample surface, by virtue of the primary electron beam, are accelerated towards the positively biased metallic ring (typically +30 to +550V) Gaseous Secondary Electron Detector (GSED). As these electrons accelerate towards the ring they undergo ionizing collisions with gas molecules producing positive ions and additional electrons known as environmental secondary electrons. The environmental electrons further ionize the gas on their way to the ring producing a cascade amplification of the original signal. The amplified signal induced in the ring is used to form an image. The electric field generated between the GSED ring and the grounded stage causes the positive ions produced in the cascade to drift towards the sample, effectively neutralizing negative charge build up on the surface of a non-conducting sample.

Secondary electron (SE) imaging in a variable pressure SEM has necessitated the development of new detector technologies. One approach is to measure the charge, Q, induced on a positive electrode placed at some distance, r, from the specimen stage which is a ground potential. Gas ionization by sufficiently energetic SEs produces electron-ion pairs which are charge separated by the applied electric field. Electrons drift towards the anode causing further gas ionization in a cascade process, and the ions drift towards the stage (cathode). The total Q induced at the anode will include charge components induced by the motion of both electrons, Q E , and ions, Q I . From Ramo's theorem [1], the charge induced on an electrode by a point charge moving in an electric field will be proportional to the ratio of the distance traveled by the charge to r. The motion of the electron and the ion will induce at total charge of ∆Q T = Q E + Q I = qx E r -1 + q(1-x E )r -1 = qx E r -1 + qx I r -1 , where q is the electronic charge, x E is the distance the electron is displaced and x I is distance the ion is displaced. The total induced current, I T , can be obtained from I T = dQ T /dt = q(x E /dt)r -1 + q(x I /dt)r -1 = qv E r -1 + qv I r -1 , where v E and v I are the electron and ion drift velocity respectively. Since the v E is around three orders of magnitude greater that v I the total induced current pulse for each electron-ion pair consists of a sharp electron peak followed by a (much weaker) slow ion tail. The maximum width of this current pulse is determined by the time taken for the ions to traverse x I which is typically 1 -100 µs, depending on r and v I . Consequently, at fast scan speeds the current induced by ions is not completely collected during the pixel dwell time.This problem has been addressed in detectors employed in nuclear physics by the use of a Frisch grid which is placed a distance, z, in front of the anode [2]. The grounded grid in this application serves two purposes. First, it electrostatically screens the anode from the movement of ions and electrons between the grid and the cathode, and second, it enables the electrons that cross the gap between the grid and the anode to induce their total charge without the involvement of the ions. By placing a Frisch grid below the anode in a VPSEM (ideally at a distance close to the gas ionization path length), the total induced current can in principle be collected at high scan rates without ionrelated artifacts, such as SE emission quenching and gas gain damping. SE gas gain below the grid can be achieved by biasing the specimen stage with a negative voltage to accelerate SEs towards the grid and ions towards the stage. The grounded grid also provides a pathway for ion recombination and efficiently removes any additional ions produced between the grid and the anode [3].A grounded Frisch grid with a 1 mm spacing was placed between the anode and the sample stage in a FEI Quanta 200. A TV rate SE image of a uncoated sapphire specimen (pixel dwell time = 100 ns) u...

The bandwidth and contrast of secondary electron (SE) images obtained using variable pressure scanning electron microscopy are enhanced when a grounded Frisch grid is placed between the SE detecting anode and the negatively biased stage. The improvement in SE image quality occurs as a consequence of the grounded Frisch grid electrostatically screening the ‘slow’ induced ion current signal, generated below the grid, from the induced current detected above the grid by the anode. Ion induced artefacts, such as image smearing at fast scan rates, are virtually eliminated using a Frisch grid. Gas amplification data are presented to illustrate that gas gain can be optimized by varying the Frisch grid–stage (amplification region) separation Frisch grid–anode (drift region) separation and stage bias.