Laser Voltage Probing (LVP) is an essential Failure Analysis (FA) technique that has been widely adopted by the industry. Waveforms that are collected allow for the analyst to understand various internal failure modes related to timing or abnormal circuit behavior. As technology nodes shrink to the point where multiple transistors reside within the diffraction-limited laser spot size, interpretation of the waveforms can become extremely difficult. In this paper we discuss some of the evolving challenges faced by LVP and propose a new technique known as Differential LVP (dLVP) that can be used to debug marginal failing devices that exhibit a pass/fail boundary in their shmoo plot. We demonstrate how separate pass and fail LVP waveforms can be collected simultaneously and compared to immediately identify whether logic is corrupted and when the corruption occurs. The benefits of this new technique are many. They include guarantees of equivalent pass vs. fail data independent of crosstalk, system noise, stage drift, probe placement, temperature effects, or the diffraction-limited resolution of the probe system. Implementing dLVP into existing tools could extend their effective lifetimes and improve their efficacy related to the demands posed by the debug of 5nm technologies and smaller geometries. We anticipate that fully integrated and evolved dLVP will complement workhorse FA applications such as Laser Assisted Device Alteration (LADA) and Soft Defect Localization (SDL) analysis. Wherein those techniques map timing marginalities propagating to, and observed by, a capture flop, dLVP can extend such capabilities by identifying the first instance of corrupted logic inside the flop and map the corruption all the way to the chip output pin.

This presentation is an application-oriented tutorial on laser-assisted device alteration (LADA) and soft defect localization (SDL) techniques and how they are used to analyze marginal digital failures and identify analog circuits that are sensitive to voltage perturbations. The presentation includes well-illustrated instructions for equipment setup and validation, guidelines for collecting and analyzing images, and examples of how to interpret pass/fail sites and assess the effect of laser interactions on circuit behaviors. It also includes a brief overview of time-resolved LADA and introduces the concept of laser-induced fault isolation (LIFA).

This presentation is an application oriented tutorial on laser-assisted device alteration (LADA) and soft defect localization (SDL) techniques and how they are used to analyze marginal digital failures and identify analog circuits that are sensitive to voltage perturbations. The presentation includes well-illustrated instructions for equipment setup and validation, guidelines for collecting and analyzing images, and examples of how to interpret pass/fail sites and assess the effect of laser interactions on circuit behaviors. It also includes a brief overview of time-resolved LADA and introduces the concept of laser-induced fault isolation (LIFA).

As advanced silicon-on-insulator (SOI) technology becomes a more widespread technology offering, failure analysis approaches should be adapted to new device structures. We review two nanoprobing case studies of advanced SOI technology, detailing the electrical characterization of a compound gate-to-drain defect as well as the characterization of unexpected SOI source-to-well leakage.

We present the first experimental demonstration of stuck-at scan chain fault isolation through the exploitation of Single Event Upsets (SEU) in a Laser-Induced Fault Analysis (LIFA) system. By observing a pass/fail flag, we can spatially map all flops after a defect in a failing scan chain through induced SEU sites produced by a fiber-amplified 25 ps 1064 nm diode laser. In addition, a custom fault isolation methodology is presented in which the result highlights only the first working flop immediately after the defect mechanism causing the stuck-at chain failure. This work demonstrates a novel method for rapid scan chain fault isolation that significantly improves localization efficacy over conventional best-known methods (BKM) based on frequency mapping. Moreover, experimental results are presented to demonstrate that LIFA can be extended to interrogate the data state of flip flops in a scan chain. Results are also presented to establish that LIFA can be configured as a hardware-based diagnostics platform.

Low power mode current is a very important parameter of most microcontrollers. A non-production prototype microcontroller had high current issues with certain SRAM modules which were produced using a new memory compiler. All devices were measuring 100’s μA of low power mode current which was an order of magnitude higher than the requirement. Many failure analysis (FA) techniques had to be used to determine the root cause: Optical Beam Induced Resistance Change (OBIRCh), photo emission microscopy (PEM), microprobing, and nanoprobe device characterization. Transistor models and measurements of probe structures from the effected lots both predicted that the device low power mode current would meet expectations; however, all first silicon samples had elevated low power mode current. A knowledge of low power design methodology was needed to ensure all issues were discovered.

We present an upgraded time-resolved LADA system, with a 25ps pulsed laser, integrated into a commercial laser scanning microscope used in failure analysis. We demonstrate the use of this system on 14nm/16nm finfet devices.

An asynchronous Low Voltage Detect (LVD) interrupt during the self-test portion of the reset sequence of a microcontroller randomly caused a corrupted clock state that was not recoverable except through a power on reset, or POR. This paper discusses the techniques used to overcome the many obstacles encountered to determine the root cause of the race condition that corrupted the clock state machine registers.

Laser-assisted device alteration (LADA) is an established technique used to identify critical speed paths in integrated circuit. In this paper, the characterization of continuous wave 1340nm laser induced currents and the LADA failure rate show that a two photon absorption explanation for the LADA effect is not plausible. The following sections confirm the results of a 28nm-node nMOS transistor using a 2.45NA solid immersion lens. The effects of global heating to that of local laser heating are then compared. The paper shows that the LADA response time to approximately 1300nm irradiation is << 100ns. It explains LADA at approximately 1300nm, free carrier absorption in the silicon and in the local silicide layers, and presents selected 1320nm LADA images on 28nm-node devices. Finally, it shows 1064nm LADA images on the same structure that indicate that 1064nm interaction with transistors is related to free carrier absorption, rather than electron-hole pair creation.

Laser-assisted device alteration (LADA) is an established technique used to identify critical speed paths in integrated circuits. LADA can reveal the physical location of a speed path, but not the timing of the speed path. This paper describes the root cause analysis benefits of 1064nm time resolved LADA (TR-LADA) with a picosecond laser. It shows several examples of how picosecond TR-LADA has complemented the existing fault isolation toolset and has allowed for quicker resolution of design and manufacturing issues. The paper explains how TR-LADA increases the LADA localization resolution by eliminating the well interaction, provides the timing of the event detected by LADA, indicates the propagation direction of the critical signals detected by LADA, allows the analyst to infer the logic values of the critical signals, and separates multiple interactions occurring at the same site for better understanding of the critical signals.

In this paper, we describe improved hardware to connect a semiconductor tester or applications board to a laser scanning microscope (LSM) for performing dynamic laser stimulation (DLS). The hardware, called DXGlue, simplifies the DLS workflow and enables new applications. We describe its precise monitoring of the fail rate and fail mode, its use for time resolved DLS and the enabling of long test loops with short laser dwell times.

As semiconductor geometries decrease, the size of a defect that leads to circuit failure also decreases. While many defects will cause photoemission or observable leakage paths, occasionally a defect will occur in an area that cannot be easily analyzed. In this analysis, a yield issue in nickel-silicide (NiSi) piping is investigated. The failure had characteristics that fell into areas that avoided detection. A planar transmission electron microscope of the substrate at the defect site was performed to look for evidence of crystalline defects that would allow a conduction path across the channel. This analysis found that NiSi encroachment was the root cause of the yield issue. All analyzed units had the defect between stacked nFET transistors. Because the defect was between stacked nFET gates, the results show that the failure characterization required control of multiple gates to measure the transistor off-state drain to source current.

Advanced technologies with higher gate leakage due to oxide tunneling current enable detection of high resistance faults to gate nodes using a straight forward resistance measurement.

We describe a technique that is used to obtain timing information from laser assisted device alteration (LADA). The technique uses a non-pulsed laser scanning microscope to obtain timing information with a temporal resolution on the order of microseconds. Custom software is used to extract the timing information from the LADA images.