Previous Page  18 / 66 Next Page
Information
Show Menu
Previous Page 18 / 66 Next Page
Page Background

edfas.org

ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 18 NO. 4

18

Fig. 3a) that is kept below the superconducting criti-

cal temperature of a few degrees Kelvin by using a

closed‑cycle cryostat (Fig. 3b). The detector is biased at

a constant current. When a photon hits the meander, it

creates a hot spot (Fig. 3c, top) that makes the current

crowd toward the remaining portion of the nanowire

(Fig. 3c, right), causing it to exceed the critical current

density value,

j

c

(Fig. 3c, bottom). At this point, the

meander superconductivity is broken (i.e., it becomes

resistive), and a voltage pulse across the SSPD (Fig. 3c, left)

can be detected by the front‑end electronics. After a few

nanoseconds, the meander returns to superconductivity

and is ready to detect a new incoming photon.

LOW-VOLTAGE SENSITIVITY

The typical experimental setup used to acquire TRE

waveforms with the SSPD is shown in Fig. 4. A pulse gen-

erator provides a clock to the DUT while its spontaneous

emission is collected by the high-NA solid immersion lens

of amicroscope and, through a single-mode fiber, is fed to

the SSPD. The delay between each detected photon from

theSSPDand the clock synchronization signal ismeasured

by the timing electronics. The cumulative histogramof all

the photon arrival times is used to reconstruct the TRE

waveform. Note that most of the optical tools (e.g., those

from FEI and Hamamatsu) can be retrofitted to perform

TRE measurements. The only requirements are a fiber

port to connect the SSPD as well as timing electronics to

reconstruct the measured waveform.

Figure 5 shows the emission intensity measured at

different chip supply voltages from an inverter gate in

32 nm silicon-on-insulator (SOI) technology using two

generations of SSPD.

[10]

The switching emission signal

is measured as the amplitude of the switching emission

peak in the TREwaveform, while the noise is the standard

deviation of the intensity level correspondent to the

semiperiod during which the field-effect transistor (FET)

is conductive. Figure 5 shows that themeasurement noise

is limited by the detector noise (dark-count rate, or DCR)

for voltages lower than 0.65 V. The second-generation

SSPD is characterized by a higher system detection effi-

ciency; as a consequence, both the switching emission

Fig. 3

(a) NbNmeander, shaped in a 9-µm-diameter circle, that was used for this work.

[10]

(b) Three-stage closed‑cycle cryostat

that houses the SSPD.

[10]

(c) Working principle of the SSPD. The detector is biased at a constant current lower than the

critical current. When a photon hits the nanowire, a hotspot is created. The current crowds at the edges, and when it

becomes higher than the critical current value, the superconductivity is lost, leading to a voltage pulse that can be

detected by external electronics.

(a)

(b)

(c)

Fig. 4

Experimental setup used to acquire TRE waveforms