Con en s lis s a ailable a ScienceDi ec
Nuclea Ins . and Me hods in Physics Resea ch, A
jou nal homepage: www.else ie .com/loca e/nima
Full Leng h A icle
nLGAD gain esponse o low-pene a ing pa icles
Jai o Villegas a,∗, Ca men To es b, Milos Manojlo ic a, M. Ca men Jimenez-Ramos b,c,
Neil Mo a a, Ja ie Ga cia Lopez b,d, Sal ado Hidalgo a
aIns i u o de Mic oelec onica de Ba celona, IMB-CNM-CSIC, 08193 Ce danyola del Valles, Ba celona, Spain
bCen o Nacional de Acele ado es, CNA, 41092 Se illa, Spain
cDepa amen o de Fisica Aplicada II, Uni e sidad de Se illa, 41012 Se illa, Spain
dDepa amen o de Fisica A omica, Molecula y Nuclea , Uni e sidad de Se illa, 41012 Se illa, Spain
ARTICLE INFO
Keywo ds:
LGAD
X- ays and cha ged pa icle de ec o s
IBIC
ABSTRACT
De ec ing low-pene a ing pa icles is undamen al o a wide ange o applica ions in he ields o indus y,
medicine, and pu e esea ch. The Low Gain A alanche De ec o buil on n- ype subs a es (nLGAD) is
conside ed a good candida e o ha , as i showed o ha e he po en ial o de ec such pa icles wi h
high sensi i i y while a oiding he high noise le els associa ed wi h a adi ional a alanche pho ode ec o .
An o e iew o he elec ical cha ac e iza ion o he la es nLGAD ba ch ab ica ed a he Ins i u e o
Mic oelec onics o Ba celona is p esen ed in his wo k. Addi ionally, gain esponse measu emen s o ligh
wi h a ying pene a ion dep hs in silicon and 600 keV p o ons a e also p esen ed. The esul s con i m he
a o emen ioned po en ial o nLGADs o de ec low-pene a ing pa icles in silicon. Addi ionally, i was obse ed
ha he gain esponse o he de ices is no only dependen on he pene a ion dep h o he s udied pa icles,
bu also on he na u e and lux o he sou ce.
1. In oduc ion
Low Gain A alanche De ec o s (LGAD) we e i s concep ualized
and ab ica ed a he Ins i u e o Mic oelec onics o Ba celona (IMB-
CNM) [1]. As hey we e o iginally hough o iming applica ions
o high-ene gy physics expe imen s, hei design and manu ac u ing
was a ge ed o de ec cha ged pa icles wi h a high ange in silicon
(mips) [2]. In o he wo ds, a good gain esponse o a adi ional IMB-
CNM LGAD is limi ed o high pene a ing pa icles in silicon, and
hei o e all pe o mance was ound o be d as ically educed when
de ec ing low-pene a ing ones. This e ec o igina es om he dep h
a which a high- o low-pene a ing pa icle deposi s i s ene gy wi hin
he de ec o [3].
In o de o o e come his disad an age o applica ions equi ing
he de ec ion o low-pene a ing pa icles, he nLGAD was de eloped
a IMB-CNM [3,4] as an e olu ion o he LGAD used o high ene gy
physics expe imen s, and based on he n- ype A alanche Pho ode ec o
o low-pene a ing pho ons [5].
The unc ionali y o an nLGAD is alike o ha o a adi ional LGAD.
Tha is, he cha ge mul iplica ion phenomenon is achie ed by di using
a mode a ely doped mul iplica ion laye in be ween an elec ode and
a high esis i i y subs a e, which c ea es a e y localized high elec ic
ield a ound he PN junc ion when he de ec o is biased. The main
di e ence o a nLGAD wi h espec o an LGAD is he conduc i i y
∗Co esponding au ho .
E-mail add ess: [email p o ec ed] (J. Villegas).
ype o he subs a e and he mul iplica ion laye , which a e n- ype
ins ead o p- ype, as well as he conduc i i y ype o he elec odes.
This has ce ain implica ions, wi h he key poin being he change
in d i di ec ion o elec ons and holes. The i s IMB-CNM nLGAD
p o o ypes demons a ed a gain o 15–20 o low-pene a ing isible
pho ons (wa eleng h o 404 nm) [3], while he gain esponse o ypical
IMB-CNM LGADs o such pa icles was below 2. An ex ensi e e iew o
he physical mechanisms ha make he cons uc ion o nLGADs op imal
o low-pene a ing pa icles can be ound in [3].
The nLGAD is designed o applica ions equi ing enhanced sensi-
i i y o low-pene a ing pa icles in silicon, such as so X- ays o UV
pho ons. While some s a egies ocus on educing he en ance window
in p- ype LGADs [6] o s anda d n- ype PiN de ec o s [7], he nLGAD
can also enhance he signal ampli ude h ough a alanche mul iplica-
ion, po en ially achie ing a gain o 15–20 o hese low-pene a ing
pa icles.
2. Fab ica ion and elec ical cha ac e iza ion o he s udied nL-
GADs
The s udied de ices we e ab ica ed on 100 mm n- ype high e-
sis i i y wa e s (CNM-4nLG1 echnology), which co esponds o he
h ps://doi.o g/10.1016/j.nima.2025.170208
Recei ed 3 Augus 2024; Recei ed in e ised o m 23 No embe 2024; Accep ed 6 Janua y 2025
Nuclea Ins umen s and Me hods in Physics Resea ch A 1072 (2025) 170208
A ailable online 15 Janua y 2025
0168-9002/© 2025 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC license ( h p://c ea i ecommons.o g/licenses/by-
nc/4.0/ ).
J. Villegas e al.
Fig. 1. Schema ics o he on iew and c oss sec ion o he s udied nLGADs.
Fig. 2. Doping concen a ion a ound he PN junc ion o he s udied nLGADs, ex ac ed
ia SRP echnique [8].
second ba ch o IMB-CNM manu ac u ed nLGADs, being he i s one
comp ehensi ely s udied in [3]. The de ices ha e an ac i e a ea and
hickness o 1.3 ×1.3 mm2and 270 μm, espec i ely (Fig. 1).
Wi hin he ac i e a ea, he n- ype mul iplica ion egion has an a ea
o 1×1 mm2. Du ing he ab ica ion, a ci cula en ance window o
0.7 mm diame e is e ched on he me al su ace co e ing he ac i e
a ea, o a oid ligh e lec ion du ing gain esponse measu emen s. The
su ace o he en ance window is composed o a laye o 0.7 μm o
silicon ni ide o e a0.4 μm laye o silicon dioxide. An nPiN wa e
was also ab ica ed in his ba ch, so i s diodes ha e he same geome y
as he one depic ed in Fig. 1.
The ne doping concen a ion o he PN junc ion composing he
nLGADs was in es iga ed ia Sp eading Resis ance P o iling (SRP)
echnique (Fig. 2). An o e iew o his e e se enginee ing echnique
can be ound in [8]. Such measu emen s allowed us o measu e he
hickness o he 𝑃++ laye (∼1 μm), which ne e ge s deple ed and
hence composes, along wi h he passi a ion laye s, he dead en ance
window o he de ice.
Capaci ance s. bias ol age measu emen s we e pe o med on mo e
han 20 de ices dis ibu ed ac oss di e en places on he wa e . The
CV cu es we e ob ained wi h a Keysigh Agilen 4284 A LCR-me e
in pa allel mode, a 20 ◦C, 10 kHz and 500 mV AC. Fig. 3displays
he a e age CV cu e o such measu emen s, showing a e y uni o m
deple ion ol age o he gain laye (𝑉𝑔 𝑙≃ 28 V) ac oss he wa e .
Alongside, leakage cu en s. bias ol age measu emen s we e ca ied
ou o he same samples, a 20 ◦C and wi h a Kei hley 2410. The
backside o he de ec o s was g ounded and he cu en was ead
om he on side. The gua d ing (Fig. 1) was connec ed in e e y
measu emen o sepa a e he su ace cu en om he leakage cu en
in he de ec o bulk. The a e age leakage cu en cu e is shown in
Fig. 3, highligh ing a leakage cu en (in he ope a ional ol age egion)
o he o de o he nA, and a mean b eakdown ol age o 𝑉𝐵 𝐷≃ 225 V.
3. Gain esponse o ul a iole , isible and in a ed ligh
Th ee nLGAD de ec o s we e in es iga ed o hei gain esponse
when exposed o ul a iole (UV) ligh o 369 nm, isible ligh o
404 nm and in a ed (IR) ligh o 1064 nm. The measu emen s we e
ca ied ou a oom empe a u e wi h a T ansien Cu en Technique
(TCT) se up [9]. The lase beam was ocused a ound he cen e o he
ac i e a ea (Fig. 1), a no mal incidence wi h espec o i and wi h
a p ojec ed a ea o ≈(𝜋∕4)60 ⋅60 μm2on he su ace o he en ance
window (o ≈(𝜋∕4)700 ⋅700 μm2). Such cen e was ound by scanning
he signal esponse o he de ices in X and Y using s eps o 2 μm. The
lase pulse equency (1 kHz) was kep cons an o all wa eleng hs and
bo h nLGAD and e e ence nPiN du ing he measu emen s. The pulse
wid h was also ixed o he minimum alue o he lase diodes, hough
i di e ed o each wa eleng h: ≈100 ps o 369 nm and ≈300 ps o
404 and 1064 nm.
The lase -gene a ed signal is ead om he de ice on isde (as
shown in Fig. 1), goes h ough a C2-HV B oadband Ci idec ampli ie ,
wi h an analog bandwid h o 2 GHz, 40 dB gain and an in eg a ed
Bias-Tee, be o e being analyzed using a DRS4 E alua ion Boa d Os-
cilloscope. Fo e e y de ice, wa eleng h and ol age poin , 10 000
wa e o ms we e a e aged o ex ac he gain, ha was in e ed by
di iding he in eg a ed ampli ied ou pu signal (Fig. 4) o he nLGADs
by ha o a nPiN o he same ab ica ion ba ch.
As he 369 nm UV ligh is he leas pene a ing in silicon o all
he s udied wa eleng hs [10], he lase in ensi y was i s con olled
o achie e an accep able Signal o Noise Ra io (SNR) in bo h he
e e ence nPiN and he nLGAD. I was ound ha a easonable SNR
o he e e ence PiN was only ob ained wi h he lase diode powe
a i s maximum alue (5 mW), o which SNR anged be ween 2 and
4 (Fig. 4) wi hin he s udied ol age ange. Fo lowe alues o he
lase diode powe , he UV signals in he nPiN we e deg aded (SNR <
2) o e en los (SNR ≤1), which would ha e impeded a gain esponse
analysis. This maximum lase diode powe alue was main ained when
e alua ing he nLGADs gain o ligh pulses o 404 and 1064 nm.
Fo hese wa eleng hs, he maximum powe alues we e 50 mW and
100 mW, espec i ely.
Ne e heless, he numbe o pho ons being abso bed wi hin he
ac i e olume o he de ec o was always smalle han hose composing
he ac ual lase pulse, ega dless o he used wa eleng h. This can be
unde s ood by e alua ing he e e ence PiN signal (Fig. 4) and he
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
2
J. Villegas e al.
Fig. 3. A e age cu es (wi h s anda d de ia ion e o s) o he capaci ance and leakage cu en measu emen s a 20 ◦C. In he a e age IV cu e, he ba e o s a e only displayed
up o he mean b eakdown ol age o 𝑉𝐵 𝐷≃ 225 V, as he ela i e e o in he cu en measu emen s g ows exponen ially a e such bias poin .
Fig. 4. TCT ou pu signals o illumina ion o an nLGAD and an nPiN wi h ligh o 369, 404 and 1064 nm wa eleng h. The Y axis is adjus ed o e e y wa eleng h o be e
dis inguished he di e ences be ween he nLGAD and he nPiN signals.
pene a ion dep h in silicon o e e y pho on species. Fo IR ligh o
1064 nm, wi h a mean pene a ion dep h o ≈1000 μm [10] (>270 μm
o ac i e hickness), mos o he pho ons will c oss he de ec o wi hou
in e ac ing wi h i , and only a negligible numbe o hem will be
abso bed close o he PN junc ion, hence igge ing he a alanche
mechanism o elec ons. Fo isible ligh o 404 nm, wi h a pene a ion
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
3
J. Villegas e al.
Fig. 5. (1): Gain esponse o TCT ligh pulses o 369, 404 and 1064 nm wa eleng h o he h ee s udied nLGADs. (2): A e age gain (wi h s anda d de ia ion e o ba s) a 100,
150 and 200 V o e e se bias as a unc ion o he pene a ion dep h [10] o he pho on species.
dep h o ≈0.1 μm [10] (<270 μm), he la ges pa o he pho ons will
be abso bed wi hin he dead en ance window (passi a ion and 𝑃++
laye s) wi hou causing a de ec able signal, and jus a ew o hem will
each he su ace o he deple ion wid h o do so. The si ua ion is e en
mo e d as ic o UV ligh o 369 nm, wi h a pene a ion dep h o ≈0.01
μm [10].
As a esul , he ac ual in ensi y wi hin he ac i e olume o he
de ec o s a ied be ween wa eleng hs, despi e he lase diodes being
ope a ed a hei maximum powe . By e alua ing he nPiN signals in
Fig. 4, we may in e ha he highes in ensi y wi hin he ac i e olume
occu s o IR ligh , ollowed by isible and UV ligh .
I is wo h no icing ha he du a ion o he lase -gene a ed signals is
highe o IR han o isible o UV ligh . This is a di ec consequence o
he dep h a which he cha ge ca ie s a e gene a ed. Fo he IR lase
beam, he pho ons a e abso bed wi h he same p obabili y all along
he ac i e hickness o he nLGAD, so bo h elec ons and holes ha e
o co e a la ge dis ance o he elec odes. Mo eo e , hese p ima y
holes (wi h a smalle mobili y han elec ons) will igge he a alanche
mechanism as hey c oss he PN junc ion in hei mo ion om he bulk
o he P elec ode. In u n, his gene a es an ex a bunch o elec ons
ha delays he all o he signal e en u he as hey a e se he 270 μm
hickness o he 𝑁elec ode. In con as , isible and UV lase pulses
a e abso bed e y close o he PN junc ion. In his si ua ion, holes a e
swi ly collec ed a he P elec ode, lea ing only he as e elec ons o
con ibu e o he signal du a ion as hey d i o he 𝑁one.
The po en ial o he nLGAD is highligh ed when we examine he
gain esponse (Fig. 5) and he signal ampli ude (in con as o he
nPiN ones) o e e y one o he a o emen ioned pho on wa eleng hs.
The inne mul iplica ion mechanism o he nLGAD allows o a g ea e
SNR o all he s udied wa eleng hs. When compa ed o an nPiN,
such imp o emen o he SNR is boos ed o UV ligh , wi h he lowes
pene a ion dep h in silicon o all he pho on species in es iga ed ia
TCT.
I is signi ican o men ion ha he e o associa ed o he e e ence
PiN signals when illumina ed wi h UV ligh was qui e la ge due o hei
low SNR (Fig. 4). In u n, his caused an impac on how accu a e he
gain o such wa eleng h could be de e mined.
3.1. Gain esponse o isible ligh as a unc ion o he beam lux
The pene a ion dep h o he a ge pa icle o be de ec ed is no
he only pa ame e ha de e mines he gain esponse o an nLGAD.
E idence o gain educ ion in p- ype LGADs o high TCT beam luxes
was obse ed in [11] o IR pho ons o 1064 nm. Simila gain esponse
measu emen s as a unc ion o he beam lux o he 404 nm wa eleng h
TCT lase we e conduc ed on one nLGAD de ice (labeled as LG13 in
Fig. 6. Cha ge in eg al o he lase -gene a ed nPiN signals (CC) dependence on 𝑉𝑡ℎ
wi hin a linea ange, a 100 V and 20 ◦C. The p ojec ed a ea 𝐴≈ (𝜋∕4)155 ⋅136 μm2
was unchanged du ing he measu emen s.
Fig. 5). By de ini ion, he beam lux is exp essed as
𝐼=𝑁𝛾
𝐴(1)
whe e 𝑁𝛾is he numbe o inciden pho ons pe uni ime ( ha is, he
in ensi y) and 𝐴is he p ojec ed a ea o he lase beam on he de ice
su ace.
Wi h a TCT se up, 𝑁𝛾can be uned by adjus ing a Digi al o Analog
Con e e (DAC) h eshold 𝑉𝑡ℎ, ha con ols he amoun o powe ha
is ed o he lase diode. The smalle he 𝑉𝑡ℎ alue is, he g ea e he
numbe o pho ons composing he ou pu lase beam pulse [12]. The
e e ence nPiN de ec o was i s measu ed o ind alues o 𝑉𝑡ℎ whe e
he in eg al o he lase -gene a ed signal (e.g. he Collec ed Cha ge o
CC) was linea . Tha is, o ind a𝑉𝑡ℎ ange whe e we can assume ha
𝐶 𝐶(𝑛𝑃 𝑖𝑁) ∼𝑁𝛾∝ −𝑉𝑡ℎ. The esul s a e shown in Fig. 6.
On he o he hand, 𝐴can be adjus ed by displacing he boa d whe e
he nLGAD is moun ed in he z posi ion ( he pe pendicula o he beam
di ec ion). By mo ing he de ec o owa ds o opposi e o he opening
o he lase sou ce, he p ojec ion o he beam on he cen e o he
de ec o su ace de- ocuses, so 𝐴can g ow la ge o smalle . As in he
p e ious sec ion, 𝐴was es ima ed by assuming an ellip ical p ojec ion
o he beam on he de ec o su ace, which axes leng h a e in e ed by
scanning in X and Y in s eps o 2 μm. La ge alues o 𝐴 ansla e in o
ha ing a bigge spa ial sp ead o he inciden pho ons on o he de ec o ,
hence educing he beam lux. Fo a gi en 𝑉𝑡ℎ, i was obse ed ha he
e e ence nPiN signal did no change wi h 𝐴. In con as , he nLGAD
signal did, as Fig. 7depic s. This esul showed he i s e idence o
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
4
J. Villegas e al.
Fig. 7. Cha ge in eg al o he lase -gene a ed nLGAD and nPiN signals (CC) dependence
on he p ojec ed a ea 𝐴, a 100 V and 20 ◦C. The DAC h eshold 𝑉𝑡ℎ = 1056 mV was
unchanged du ing he measu emen s. The CC is no malized, o e e y de ice, wi h
espec o i s maximum alue ac oss 𝐴, in o de o be e dis inguish he end o he
nLGAD and he e e ence nPiN.
Table 1
DAC, CC(nPiN) a 20 ◦C and 100 V, axes leng h in X and Y o he beam p ojec ion,
es ima ed A and ela i e lux.
DAC (mV) CC(nPiN)
(a b)
X (μm) Y (μm) A=(𝜋∕4)XY
(μm2)
Rel. lux
(%)
1353 109.7 150 144 16 953 3.7
1254 143.5 162 135 17 143 4.8
1056 235.7 161 148 18 706 7.2
1056 235.7 133 122 12 801 10.5
825 318.9 144 138 15 675 11.6
1056 235.7 108 97 8285 16.2
825 318.9 115 113 10 203 17.8
1056 235.7 83 73 4775 28.2
825 318.9 91 90 6388 28.5
825 318.9 69 68 3686 49.4
1056 235.7 60 52 2415 55.7
825 318.9 48 48 1819 100
gain supp ession in nLGADs, as bo h he signal heigh and i s CC go
educed o smalle alues o 𝐴(i.e. o la ge alues o he beam lux).
Bo h he DAC h eshold 𝑉𝑡ℎ and he p ojec ed a ea 𝐴we e swep
o in es iga e he nLGAD gain esponse dependence on he 404 nm
wa eleng h beam lux, which was es ima ed o be p opo ional o
𝐶 𝐶(𝑛𝑃 𝑖𝑁 , 𝑉𝑡ℎ)∕𝐴.Table 1shows he DAC, CC(nPiN) and A alues ha
we e used du ing he measu emen s.
The measu emen s we e ca ied ou a 100 V and 20 ◦C. 𝑉𝑡ℎ was
swep wi hin he linea ange p esen ed in Fig. 6, while 𝐴was always
kep smalle han hal he a ea o he en ance window (≈(𝜋∕4)700⋅700
μm2) o ensu e ha he e was no e lec ion o he inciden pho ons wi h
he me al. The beam was always poin ed a he cen e o he en ance
window and a no mal incidence. Fo e e y lux poin , he gain was
in e ed by di iding he in eg a ed ou pu signal o he nLGAD by he
nPiN one. The esul s a e shown in Fig. 8, whe e a clea gain educ ion
wi h inc easing beam lux is obse ed. In his igu e, he lux was scaled
o i s maximum alue, which co esponds o he minimum alue o bo h
𝑉𝑡ℎ and 𝐴. This esul highligh s ha he gain esponse o an nLGAD o
low-pene a ing pho ons is subjec ed o bo h he pene a ion dep h o
he species and he beam lux o he sou ce.
4. Gain esponse o low-ene gy p o ons
Gain esponse o low-ene gy p o ons measu emen s we e pe o med
using he nuclea mic op obe beamline a he Cen o Nacional de
Acele ado es (CNA, Se ille) using a 600 keV p o on beam. P o ons
wi h such ene gy a e low-pene a ing in silicon, as hei ange is ≈10
Fig. 8. nLGAD gain esponse, a 100 V and 20 ◦C, as a unc ion o he 404 nm
wa eleng h lase beam lux (as de ined in Eq. (1)). The in ensi y is scaled o i s
maximum alue, which co esponds o he minimum alue o 𝑉𝑡ℎ and 𝐴in Table 1.
μm [13]. The mic op obe is connec ed o a 3 MV andem accele a o
ha p o ides di e en species and ion ene gies [14]. I showcases a
se o mic ome e sli s designed o a oid adia ion damage du ing Ion
Beam Induced Cu en (IBIC) cha ac e iza ion and o p e en he ion
beam halo, which dec eases he spa ial esolu ion. The quad upole lens
sys em allows o ocusing he ion beam o mic ome e dimensions. This
se up is u he complemen ed by a 2D scanning sys em synch onized
wi h he acquisi ion one, allowing spa ial in o ma ion by enabling
mapping wi h excellen spa ial esolu ion o a eas o up o a ew
millime e s. A mo e de ailed desc ip ion o he nuclea mic op obe
beamline is p o ided in [15].
IBIC measu emen s can be conduc ed wi h he ion beam inciden
a di e en angles wi h espec o he no mal o he de ec o . This is
enabled by a special sample holde capable o o a ing unde acuum
wi h an accu acy o 1 deg ee. Fu he mo e, he signal acquisi ion and
subsequen p ocessing we e ca ied ou using a con en ional elec onic
chain, comp ising an HV sou ce (model NHR 22 20X), a CANBERRA
p e-ampli ie (Model 2003BT), a Tennelec TC 245 ampli ie , and con-
nec ed o he OMDAQ acquisi ion so wa e. One nLGAD and one e -
e ence nPiN de ec o we e es ed ia IBIC. The nLGAD unde es was
p e iously measu ed ia TCT, and co esponds o he one labeled as
LG13 in Fig. 4. The e e ence nPiN was also he same as he one used
du ing he TCT measu emen s epo ed in he p e ious sec ion.
Fig. 9shows a ypical ene gy spec um ob ained when conduc ing
an IBIC 2D scan on he nLGAD unde es .
The ene gy deposi ed by a 600 keV p o on in he nLGAD sample
gene a es an elec ical signal, ha is p ocessed by he a o emen ioned
elec onics chain. The ene gy spec um can u he be imaged as a 2D
mapping, which allows o e alua e he channel ange whe e he coun s
co espond o p o ons c ossing he mul iplica ion laye (Fig. 10). Once
such channel ange is iden i ied, he collec ed cha ge (CC) in he de ice
can be de ined as i s cen oid (Fig. 9), ob ained ia Gaussian i . The
same p ocedu e is hen done wi h he e e ence nPiN, which allows us
o es ima e he gain by di iding he CC o he nLGAD by he nPiN one.
The CC expe imen al e o is s a is ically in e ed by di iding he Full
Wid h a Hal Maximum (FWHM) o he Gaussian i by he squa e oo
o numbe o coun s wi hin i s channel ange. In Fig. 10, he spec a
o he nPiN and nLGAD we e ob ained a no mal incidence, 50 V and
oom empe a u e. The 2D sweep in he nPiN case was done a ound he
ac i e a ea (disca ding pe iphe y elemen s), eason why only a peak is
obse ed in he ene gy spec um.
I is wo h no icing ha he ene gy peak a ound he nLGAD ac i e
a ea is way b oade han ha o he nPiN. This is due o s a is ical
luc ua ions in he numbe o a alanche-gene a ed cha ge ca ie s. The
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
5
J. Villegas e al.
Fig. 9. Ene gy spec um, a no mal incidence, 50 V and oom empe a u e, o 600 keV p o ons impinging on he nLGAD unde es . The le pic u e shows he spec um in
loga i hmic scale, whe e coun s in he nLGAD pe iphe y elemen s can be g asped. The igh pic u e shows, in linea scale, only he ene gy ange a ound he ac i e a ea, along
wi h i s Gaussian cen oid.
Fig. 10. (1): Spec a o he nPiN and nLGAD, a no mal incidence, 50 V and oom empe a u e. (2): Image econs uc ion o he nLGAD spec um in (1), whe e bo h he ac i e
a ea and he pe iphe y elemen s (1) can be dis inguished. The channel numbe has been scaled, in bo h pic u es, o he CC o he e e ence nPiN, so a gain spec um and map
a e ob ained.
ene gy esolu ion a 50 V, es ima ed as 𝐹 𝑊 𝐻 𝑀∕𝐶 𝐶, was ound o be
0.5% o he nPiN and 17.7% o he nLGAD.
4.1. Gain esponse s. e e se bias a no mal incidence
Wi h he a o emen ioned me hodology, gain esponse measu e-
men s we e conduc ed on he de ices unde es . The i s se o
measu emen s we e ca ied ou a oom empe a u e, no mal incidence
and sweeping he bias ol age. The gain esponse was e alua ed bo h
by impinging he 600 keV p o on beam h ough he on side (𝑃++ in
Fig. 1) and he backside (𝑁++ in Fig. 1) o he de ices. Taking in o
accoun he hickness o he passi a ion, 𝑃++ and mul iplica ion laye s
(Figs. 1and 2), a 600 keV p o on has i s B agg peak wi hin he high-
esis i i y subs a e when he beam is inciden h ough he nLGAD
on side. In pa icula , he posi ion o such peak occu s a a dep h
o ≈6.5 μm, se ing he ze o dep h a he PN junc ion, as displayed in
Fig. 2. Gi en his scena io, holes will be he main con ibu o o igge
he a alanche mechanism as hey a e d i ed o he 𝑃++, lea ing jus
a ew elec ons o do so on he opposi e di ec ion. On he o he hand,
impinging he p o on beam h ough he backside lea e us wi h holes
as he sole con ibu o o igge he a alanche mechanism. The esul s
a e shown in Fig. 11, along wi h he gain esponse o 1064 nm IR TCT
beam pulses al eady p esen ed in Fig. 5.
As in he case o ha ing he 600 keV beam inciden o he de ice
on side, mos o he a alanche-gene a ed cha ge is caused o he d i
o holes o TCT IR pho ons unde he same inciden condi ions. In u n,
Fig. 11. nLGAD gain esponse o 600 keV p o ons (wi h he beam inciden o he
de ice on and backside) and 1064 nm IR TCT beam pulses.
his would esul in a simila gain esponse o bo h pa icle species,
as con i med by he esul s in Fig. 11. I is wo h no ing ha gain
supp ession e ec s s ill need o be in es iga ed bo h o TCT IR pho ons
and 600 keV p o ons.
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
6
J. Villegas e al.
Fig. 12. (1): B agg peak dep h dependence on o a ion angle, in e ed ia SRIM simula ions, o 600 keV p o ons impinging on he s udied nLGAD. (2): nLGAD gain esponse, a
50 V and oom empe a u e, o 600 keV p o ons as a unc ion o he B agg peak dep h.
4.2. Gain esponse dependence on B agg peak dep h
The in e es in o a ing he nLGAD du ing he measu emen s is o
modi y he dep h o he B agg peak o he inciden p o ons wi hin he
de ec o . In o he wo ds, he mo e we o a e he nLGAD holde , he
close o he de ec o su ace he ions deposi he majo i y o hei
ene gy. This occu s due o he longe ack wi hin he dead en ance
window laye s ha p o ons need o a e se as he o a ion angle
inc eases. Taking in o accoun he hicknesses o he dead en ance win-
dow laye s ha co e he nLGAD mul iplica ion laye (Figs. 1and 2),
he B agg peak dep h dependence on o a ion angle o 600 keV p o ons
can be es ima ed ia SRIM simula ion [16], as shown in Fig. 12(1). The
gain esponse esul s ob ained by a ying he incidence angle (i.e. he
B agg peak dep h) a oom empe a u e a e also shown in Fig. 12(2).
The e e se bias was ixed a 50 V because, a highe ol ages and
incidence angles (co esponding o high inne gain alues), he noise
in he p e-ampli ie inc eased signi ican ly.
The analysis o he esul s e eals ha he e a e h ee gain esponse
egions based on he p o ons B agg peak dep h wi hin he de ice. When
he B agg peak is loca ed wi hin he bulk, he gain is nea ly cons an ,
and holes a e he majo con ibu o o cause impac ioniza ion. When
he B agg peak is loca ed wi hin he mul iplica ion laye , mo e elec-
ons s a o be in ol ed in he a alanche mechanism, so he gain
inc eases s eadily om ≈2 o ≈4. Las ly, once he B agg peak is loca ed
wi hin he 𝑃++ laye , he gain esponse is boos ed om ≈4 o ≈15.
On one hand, his is due o he la ge numbe o elec ons igge ing
a alanche mul iplica ion. On he o he hand, he ioniza ion cha ge
densi y wi hin he PN junc ion d as ically d ops once he B agg peak is
loca ed wi hin he 𝑃++, as he gene a ed cha ge is quickly ecombined
in his laye . In u n, his may educe po en ial gain supp ession e ec s.
4.3. Po en ial applica ions
O e all, he esul s in Fig. 12 se a s epping s one o he po en ial
use o nLGADs in applica ions equi ing e icien de ec ion o low-
pene a ing cha ged pa icles, such as neu on de ec ion sys ems [17],
plasma hea ing o nuclea usion expe imen s [18] o he s udy o he
na u al 𝛽-decay o neu ons [19,20].
Neu ons a e neu al pa icles, making hem di icul o de ec di-
ec ly wi h silicon de ec o s. Howe e , by coa ing he en ance window
wi h speci ic ma e ials, neu ons can be de ec ed indi ec ly. Such ma e-
ials include Li-6 o B-10, ha a e able o cap u e neu ons, a eac ion
ha p oduces o he cha ged pa icles, including alphas o ∼1 MeV [17].
An e icien de ec ion o hese low-pene a ing alpha pa icles, wi h
a ange in silicon o ∼1 μm [13], is essen ial in hese expe imen s.
On he o he hand, plasma hea ing o nuclea usion expe imen s is
o en pe o med by injec ion o p o ons accele a ed a ≈100 keV [18].
A p ope pa icle de ec ion sys em o hese p o ons, wi h a ange in
silicon o ∼1 μm [13] is needed o plasma diagnosis du ing i s hea ing.
Finally, he s udy o he na u al 𝛽-decay o neu ons [19,20] equi es
o de ec and analyze he p ope ies o he emnan p o on o such
eac ion, wi h an ene gy o ≈1 keV and a ange in silicon o ∼ 0.01 μm.
5. Conclusions
The s udies p esen ed in his wo k demons a e ha he nLGAD gain
esponse is enhanced o low-pene a ing pho ons de ec ion, a esul
al eady obse ed in [3,4]. Mo eo e , hey also highligh ha he lux
o he beam sou ce is a key pa ame e o conside when e alua ing
he pe o mance o such de ices, as highe luxes lead o lowe gain
alues. This is a di ec consequence o ha ing a highe ioniza ion
cha ge densi y a ound he PN junc ion (i.e. whe e he elec ic ield is
a i s maximum alue), an e ec al eady obse ed in [11].
Addi ionally, IBIC measu emen s o 600 keV p o ons sugges ha
he gain esponse is enhanced in he case o low-pene a ing cha ged
pa icles de ec ion, as long as hey deposi hei ene gy close o he
de ice su ace. As a case in poin , i has been obse ed ha ela i ely
mode a e le els o gain (5–20) a e ob ained only when elec ons a e
he majo con ibu o o igge he a alanche mechanism. The gain
supp ession mechanism may also be p esen in nLGADs when de ec ing
cha ged pa icles, as ound in [21] when e alua ing he gain esponse
o p- ype LGADs. Howe e , u he measu emen s a e equi ed o ully
con i m his hypo hesis.
O e all, he gain esponse esul s es ablish a basis o he po en ial
use o he nLGAD in applica ions ha equi e e icien de ec ion o
low-pene a ing pho ons o cha ged pa icles.
CRediT au ho ship con ibu ion s a emen
Jai o Villegas: W i ing – e iew & edi ing, W i ing – o iginal
d a , Me hodology, In es iga ion, Fo mal analysis, Da a cu a ion. Ca -
men To es: Me hodology, Fo mal analysis, Da a cu a ion. Milos
Manojlo ic: Fo mal analysis, Da a cu a ion. M. Ca men Jimenez-
Ramos: Valida ion, Supe ision, P ojec adminis a ion, Me hodol-
ogy, In es iga ion, Funding acquisi ion, Fo mal analysis. Neil Mo a :
Me hodology, In es iga ion. Ja ie Ga cia Lopez: Valida ion, Supe -
ision, P ojec adminis a ion, Me hodology, In es iga ion, Funding
acquisi ion, Fo mal analysis. Sal ado Hidalgo: Valida ion, Supe -
ision, P ojec adminis a ion, Me hodology, In es iga ion, Funding
acquisi ion, Concep ualiza ion.
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
7
J. Villegas e al.
Decla a ion o compe ing in e es
The au ho s decla e ha hey ha e no known compe ing inan-
cial in e es s o pe sonal ela ionships ha could ha e appea ed o
in luence he wo k epo ed in his pape .
Acknowledgmen s
This wo k has been unded by he Spanish Minis y o Science,
Inno a ion and Uni e si ies (MICIU/AEI/10.13039/501100011033/)
and by he Eu opean Union’s ERDF p og am ‘‘A way o making Eu-
ope’’. G an e e ences: PID2020-113705RB-C32, PID2021-124660OB-
C22, PDC2021-121718-C32 and PDC2023-145925-C32. Also, i was
suppo ed by he ollowing Eu opean unding p og ams: Eu opean
Union’s Ho izon 2020 Resea ch and Inno a ion (unde G an Ag ee-
men No. 101004761, AIDAInno a) and Nex Gene a ionEU (PRTR-
C17.I1).
M. Ca men Jimenez-Ramos acknowledges he suppo o his wo k
h ough a VI PPIT-US con ac . This in es iga ion has also been pa -
ially inanced by he p ojec e . ASTRO21/1.1/1 wi h inancing om
he Eu opean Union- Nex Gene a ionEU, he Minis y o Science, Inno-
a ion and Uni e si ies, Reco e y Plan, T ans o ma ion and Resilience,
Spain, he Depa men o Uni e si y, Resea ch and Inno a ion o he
Jun a de Andalucia and he Uni e si y o Se illa, Spain.
Da a a ailabili y
Da a will be made a ailable on eques .
Re e ences
[1] G. Pelleg ini, e al., Technology de elopmen s and i s measu emen s o low gain
a alanche de ec o s (LGAD) o high ene gy physics applica ions, Nucl. Ins um.
Me hods Phys. Res. A 765 (2014) 12–16, h p://dx.doi.o g/10.1016/j.nima.2014.
06.008.
[2] C. G ieco, e al., O e iew o CNM LGAD esul s: Bo on Si-on-Si and epi axial
wa e s, J. Ins um. 17 (C09021) (2022) h p://dx.doi.o g/10.1088/1748-0221/
17/09/C09021.
[3] J. Villegas, e al., Gain measu emen s on NLGAD de ec o s, Nucl. Ins um.
Me hods Phys. Res. A 1055 (168377) (2023) h p://dx.doi.o g/10.1016/j.nima.
2023.168377.
[4] W. Khaled, e al., Fi s esul s o he pLGAD senso o low-pene a ing pa icles,
Nucl. Ins um. Me hods Phys. Res. A 1040 (167220) (2022) h p://dx.doi.o g/
10.1016/j.nima.2022.167220.
[5] Hamama su Si APD ca alog, 2024, [Online]. A ailable: h ps://www.hamama su.
com/con en /dam/hamama su-pho onics/si es/documen s/99_SALES_LIBRARY/
ssd/si_apd_kapd0001e.pd .
[6] V. Hinge , e al., Resol ing so X- ay pho ons wi h a high- a e hyb id pixel
de ec o , F on . Phys. 12 (2024) h p://dx.doi.o g/10.3389/ phy.2024.1352134.
[7] J. Segal, e al., Thin-en ance window p ocess o so X-Ray senso s, F on . Phys.
9 (2021) h p://dx.doi.o g/10.3389/ phy.2021.618390.
[8] R.G. Mazu , D.H. Dickey, A sp eading esis ance echnique o esis i i y mea-
su emen s on silicon, J. Elec ochem. Soc. 113 (255) (1966) h p://dx.doi.o g/
10.1149/1.2423927.
[9] G. K ambe ge , Ad anced ansien cu en echnique sys ems, in: The 23 d
In e na ional Wo kshop on Ve ex De ec o s, 2015, h p://dx.doi.o g/10.22323/
1.227.0032.
[10] M.A. G een, Sel -consis en op ical pa ame e s o in insic silicon a 300 K
including empe a u e coe icien s, Sol. Ene gy Ma e . Sol. Cells 92 (4) (2008)
1305–1310, h p://dx.doi.o g/10.1016/j.solma .2008.06.009.
[11] E. Cu as, e al., Gain supp ession mechanism obse ed in low gain a alanche
de ec o s, Nucl. Ins um. Me hods Phys. Res. A 1031 (166530) (2022) h p:
//dx.doi.o g/10.1016/j.nima.2022.166530.
[12] Pa icula s lase use manual, e sion 2.0, 2015, [Online]. A ailable: h ps:
//www.pa icula s.si/index.php.
[13] M. Be ge , e al., S opping-powe and ange ables o elec ons, p o ons, and
helium ions, NISTIR (4999) (2017) h p://dx.doi.o g/10.18434/T4NC7P.
[14] J. Gomez-Camacho, e al., Resea ch acili ies and highligh s a he cen o
nacional de acele ado es (CNA), Eu . Phys. J. Plus 136 (273) (2021) h p:
//dx.doi.o g/10.1140/epjp/s13360-021-01253-x.
[15] M.A. On alba, e al., Ex e nal mic obeam se -up a he CNA (se illa) and i s
applica ion o he s udy o a esic jewelle y, Nucl. Ins um. Me hods Phys. Res.
B 181 (2001) 664–669, h p://dx.doi.o g/10.1016/S0168-583X(01)00355-X.
[16] J.F. Ziegle , e al., SRIM – he s opping and ange o ions in ma e , Nucl.
Ins um. Me hods Phys. Res. B 268 (11–12) (2010) 1818.–1823, h p://dx.doi.
o g/10.1016/j.nimb.2010.02.091.
[17] B.F. Phlips, e al., Neu on de ec ion using la ge a ea silicon de ec o s, Nucl.
Ins um. Me hods Phys. Res. A 579 (1) (2007) 173–176, h p://dx.doi.o g/10.
1016/j.nima.2007.04.033.
[18] H. Han, e al., A sus ained high- empe a u e usion plasma egime acili a ed by
as ions, Na u e 609 (2022) 269—-275, h p://dx.doi.o g/10.1038/s41586-022-
05008-1.
[19] L.J. B oussa d, e al., Using nab o de e mine co ela ions in unpola ized neu on
decay, Hype ine In e ac . 240 (1) (2019) h p://dx.doi.o g/10.1007/s10751-
018-1538-7.
[20] D. Mose , e al., NoMoS: An R×B d i momen um spec ome e o be a decay
s udies, in: In e na ional Wo kshop on Pa icle Physics a Neu on Sou ces, PPNS
2018, no. 04003, 2018, h p://dx.doi.o g/10.1051/epjcon /201921904003.
[21] M.C. Jimenez, e al., S udy o ioniza ion cha ge densi y-induced gain supp ession
in LGADs, Senso s 22 (3) (2022) h p://dx.doi.o g/10.3390/s22031080.
Nuclea Ins . and Me hods in Physics Resea ch, A 1072 (2025) 170208
8
