Sac

JAa

b

c

d

ARRAA

P88867

KGNPC

1

uiciiSfiNtsm

S

0h

Applied Surface Science 280 (2013) 632– 640

Contents lists available at SciVerse ScienceDirect

Applied Surface Science

jou rn al h omepa g e: www.elsev ier .com/ locate /apsusc

hadowed off-axis production of Ge nanoparticles in Ar gastmosphere by pulsed laser deposition: Morphological, structural andharge trapping properties

. Martín-Sáncheza,b,∗, I. Capanc, A. Chahbouna,d, S.R.C. Pintoa, E.M.F. Vieiraa,.G. Roloa, M.J.M. Gomesa

University of Minho, Centre of Physics and Physics Department, Braga 4710-057, PortugalLaser Processing Group, Instituto de Óptica, CSIC, C/ Serrano 121, 28006 Madrid, SpainRudjer Boskovic Institute, Bijenicka 54, 10 000 Zagreb, CroatiaFST Tanger, Physics Department, BP 416 Tanger, Morocco

a r t i c l e i n f o

rticle history:eceived 14 November 2012eceived in revised form 25 April 2013ccepted 30 April 2013vailable online 18 May 2013

ACS:1.05.Hd1.07.Bc1.15.Fg1.46.Hk

a b s t r a c t

In this work, a novel customized shadowed off-axis deposition set-up is used to perform an original studyof Ge nanoparticles (NPs) formation in an inert Ar gas atmosphere by pulsed laser deposition at roomtemperature varying systematically the background Ar gas pressure (Pbase(Ar)), target–substrate distance(d) and laser repetition rate (f). The influence of these parameters on the final NPs size distributions isinvestigated and a fairly uniform droplets-free and non-agglomerated NPs distribution with averageheight 〈h〉 = 2.8 ± 0.6 nm is obtained for optimized experimental conditions (Pbase(Ar) = 1 mbar; d = 3 cm;f = 10 Hz) with a fine control in the NPs density (from 3.2 × 109 cm−2 to 1.1 × 1011 cm−2). The crystallinequality of as-deposited NPs investigations demonstrate a strong dependence with the Ar gas pressure anda crystalline to amorphous phase volume fraction �c > 50% is found for Pbase(Ar) = 2 mbar. The NPs func-tionality for charge trapping applications has been successfully demonstrated by capacitance–voltage

3.61.Le

eywords:ermaniumanostructuresulsed laser deposition

(C–V) electrical measurements.© 2013 Elsevier B.V. All rights reserved.

harge trapping

. Introduction

Semiconductor quantum dots/nanoparticles (NPs) presentnique size-dependent physical properties for many applications

n optoelectronics, non-volatile memories and solar cells [1–5]. Inomparison with Si NPs that have been widely investigated dur-ng the last two decades [3,5–9], the interest in Ge NPs has beenncreasing during the last years due to Ge advantages in respect toi such as a larger Bohr exciton radius, lower band-gap (strong con-nement), larger carrier mobility and higher band-gap sensibility toPs size variations [10]. Although a lot of work has been dedicated

o the chemical synthesis of Ge NPs with promising results [11,12],everal constraints still remain especially for applications in theicroelectronic technology because of the presence of organic

∗ Corresponding author at: Laser Processing Group, Instituto de Óptica, CSIC, C/errano 121, 28006 Madrid, Spain. Tel.: +34 915616800.

E-mail address: javier.martin.nano@gmail.com (J. Martín-Sánchez).

169-4332/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.apsusc.2013.04.170

ligands (as surfactants) and it usually requires a multi-step produc-tion process [11]. The production of Ge NPs embedded in dielectricoxide matrices has been carried out by many deposition tech-niques such as magnetron sputtering [13], ion implantation [14],thermal/electron induced evaporation [9,15] and plasma-enhancedchemical vapour deposition [16]. The NPs self-assembling in thedielectric host matrix is usually obtained through a post-depositionannealing step of a Ge-rich oxide matrix [17] or a Ge layer [18].Although this process can lead to the formation of a relatively welllayered structure of Ge NPs between dielectric oxide layers, Ge out-diffusion from the film or Ge diffusion between alternated layerscannot be ruled out even at relatively low temperatures (700 ◦C)in either Al2O3 or SiO2 matrices [19–21]. In comparison with hightemperature (800–1000 ◦C) and long (>30 min) conventional fur-nace annealing treatments, rapid thermal annealing (RTA) has been

successfully demonstrated as an effective short-time thermal treat-ment for these purposes [22,23].

In contrast to other techniques, pulsed laser deposition (PLD)is a versatile technique for the growth of high quality thin films

Surfa

acogshdestktlsitmsspsdksdlsp

GsemtoattNfccmG

2

2

ltmbHde(firNaoon

J. Martín-Sánchez et al. / Applied

nd nanostructured materials that offers the possibility of aontaminants-free synthesis process of nanoparticles (NPs) in ane-step process at room temperature (RT) when an inert/reactiveas atmosphere is introduced in the deposition chamber. The PLDynthesis of NPs in different inert and reactive gas atmospheresas been widely demonstrated mostly for Si [7], while the pro-uction of Ge NPs in background inert gas atmospheres is lessxplored [24]. Experimental and theoretical results have demon-trated that the Si NPs distributions can be changed by controllinghe experimental parameters such as background gas pressure [25],ind of background gas [26] or laser energy fluence [8]. However,here is some controversy in the results presented throughout theiterature in terms of NPs size distribution that may be due to re-puttering processes on the NPs film deposited on the substratenduced by direct deposition of droplets simultaneously depositedogether with high energetic particles [27,28], hindering the actual

orphology and distribution of NPs produced in the gas atmo-phere. For example, Yoshida et al. [25] found that the averageize of Si NPs increases monotonically with the background gasressure, whereas Lowndes et al. [29] found a maximum averageize for a critical pressure, that is attributed mainly to a differenteposition/collection geometry. On the other hand, it is also wellnown that NPs agglomerates formation produced in a gas phasetrongly depends on the experimental conditions used in the pro-uction process [30]. In this regard, for some practical applications

ike memory devices where obtaining well separated NPs on theubstrate is mandatory, it would be desirable to optimize the PLDroduction process.

In this work, we have performed a systematic study of PLDe NPs production in an inert Ar gas atmosphere using a uniquehadowed off-axis deposition geometry [31], that allows us toxplore the Ge NPs formation process in the gas phase and opti-ize the PLD parameters (background Ar gas pressure Pbase(Ar),

arget–substrate distance d and laser repetition rate f) in order tobtain functional droplets-free non-agglomerated NPs films whilevoiding the influence of high energetic particles. The influence ofhe number of laser pulses for a given experimental conditions onhe NPs density obtained on the substrate surface was studied. ThePs structural properties were evaluated on as-deposited at dif-

erent Pbase(Ar) values and post-processed by RTA samples. Theharge trapping capabilities of the Ge NPs is discussed based onapacitance–voltage (C–V) electrical measurements performed onetal-oxide-semiconductor (MOS) structures where a well definede NPs layer was embedded between thin amorphous Al2O3 layers.

. Experimental

.1. Samples fabrication

The PLD set-up used in this work is equipped with a KrF excimeraser (wavelength of 248 nm and 20 ns pulse duration) and a multi-arget carrousel assembly which allows the deposition of different

aterials by moving different targets into and out of the laseream focal point with no need to break the vacuum conditions.igh purity (99.999%) Al2O3 and Ge targets were used. Duringeposition, the targets were rotated to maintain uniform materialvaporation as much as possible. P-doped and low-resistivity Si1 0 0) substrates (1–6 � cm−1) used in all the experiments wererst dipped for 10 s in a HF:H2O (10:1) solution for native oxideemoval, rinsed in abundant deionized water and blown using dry2. Briefly, our experimental set-up consists of a mobile assembly

s shown schematically in Fig. 1, which allows on-axis and shadowedff-axis (mask height = 0.6 cm) configurations for the depositionf Al2O3 and Ge NPs, respectively. More details about the tech-ique can be found in Ref. [31]. The Al2O3 layers were deposited in

ce Science 280 (2013) 632– 640 633

high vacuum conditions (Pbase = 2 × 10−7 mbar) with d = 4 cm andf = 10 Hz using the on-axis configuration. The laser energy fluencewas kept constant in all the experiments to a value of 1.2 J/cm2. Forthe Ge NPs production, an inert Ar gas atmosphere was introducedinside the growth chamber and the Ge target was ablated using theshadowed off-axis configuration. In order to study the influence ofthe PLD experimental parameters on the Ge NPs distributions, a setof samples was prepared varying the background Ar gas pressure,target–substrate distance and laser repetition rate.

2.2. AFM and SEM characterization

The morphological characterization was performed by AFM andscanning electron microscopy (SEM) on uncapped samples of GeNPs deposited on Si (1 0 0) substrates where a 10-nm-thick amor-phous Al2O3 thin film was previously deposited. This Al2O3 layerpresents a very low root mean square (RMS) surface roughnessvalue RMS ∼ 0.15 nm as measured by atomic force microscopy(AFM) and therefore, it was used as a good reference for themeasurement of individual Ge NPs. The AFM measurements wereperformed in tapping mode using an AFM Nanoscope III set-upin ambient conditions and WSxM software was used for AFMdata analysis [32]. SEM measurements were performed with a FEIQuanta400FEG ESEM system equipped with a EDAX Genesis X4MEDS system.

2.3. Micro-Raman spectroscopy

Micro-Raman spectroscopy is a non-destructive, fast andsensitive technique for crystalline and/or amorphous phaseidentification in nanometer-sized thin films of particles. The struc-tural characterization was performed on tri-layer [Ge NPs/Al2O3(10 nm)] × 3 structures deposited on 10-nm-thick Al2O3-coatedSi (1 0 0) substrate. A Jobin-Yvon T64000 system with an opticalmicroanalysis system and a CCD detector at room temperatureusing the 488 nm Ar+ laser excitation line was used. The scatteringspectra of the samples were recorded in backscattering geometry.The laser beam was focused on the sample surface with a beamspot size of 1 �m and a power of 0.5 mW to avoid the heating ofthe sample. Silicon TO mode at 521 cm−1 was used as calibrationfrequency reference.

2.4. Capacitance–voltage (C–V) measurements

High frequency (1 MHz) C–V measurements were performed onMOS capacitor structures with Ge NPs embedded in a tri-layerstructure [Al2O3 oxide gate layer (20 nm)/Ge NPs/Al2O3 tunnellayer] at RT and 150 K with a Sula Technologies Spectrometer. If notstated the contrary, the tunnel layer thickness (ttunnel) was kept con-stant to ttunnel = 10 nm. A gold layer was deposited on the back-sidesurface of the p-type Si substrate by thermal evaporation to makean ohmic contact. Gold contacts on the front-side of the sampleswere prepared by Au thermal evaporation through a mask on topof the sample with circular openings (0.8 mm2). Possible dynamicrecharging effects during measurement were avoided using a ramprate of 0.03 V/s. The samples were kept in a cryostat during mea-surements to minimize noise from external electromagnetic fieldsand photon-induced carrier excitation.

3. Results

In Fig. 2, SEM and high resolution AFM images of Ge parti-cles distributions obtained varying the experimental parameters(number of laser pulses on Ge target np = 300 pulses) are shown.The average NPs height and their dispersion values were obtained

634 J. Martín-Sánchez et al. / Applied Surface Science 280 (2013) 632– 640

F ons fou h back

fAd

i〈ola(i(adetthrdedbrb

tPct�nw

atpa

ig. 1. Schematic representation of shadowed off-axis and on-axis (inset) configuratising shadowed off-axis configuration is obtained by scattering of produced NPs wit

rom 1 �m × 1 �m AFM images and SEM measurements (Fig. 2l–n).gglomerates of NPs were not considered in the quantitative sizeistribution analysis.

In particular, the average height of the collected Ge NPs slightlyncreases with the Ar gas pressure for Pbase(Ar) > 0.5 mbar fromh〉 = 2.8 ± 0.6 nm to 〈h〉 = 6 ± 3 nm, accompanied by the formationf agglomerates for the highest pressure (2 mbar). However, for theowest pressure Pbase(Ar) = 0.5 mbar, it is obtained the highest NPsverage height (17 ± 3 nm) with almost no agglomerates formationFig. 2a–d). On the other hand, the measured average NPs heightncreases with the target–substrate distance from 〈h〉 = 2.8 ± 0.6 nmd = 3 cm) to 〈h〉 = 6 ± 3 nm (d = 5 cm), whereas the appearance ofgglomerates is specially favored for the largest target–substrateistance (d = 5 cm) (Fig. 2e–g). The interaction dynamics of thevaporated material with Ar ambient atoms in the area betweenhe target and the substrate was studied by varying the laser repeti-ion rate for Pbase(Ar) = 1 mbar (Fig. 2h–k). A minimum average NPseight distribution with the lowest dispersion value for a criticalepetition rate fc = 10 Hz is found, whereas both average height andispersion values increase for other repetition rates. These inter-sting results clearly demonstrate that the NPs size distributionepends on the interaction degree of the evaporated material withackground gas atoms that eventually is conditioned by the laserepetition rate, in qualitative agreement with the results reportedy Wang et al. for Si NPs [33].

The results reported in Fig. 2 show that the best NPs size dis-ribution uniformity is obtained by using the PLD parameters:base(Ar) = 1 mbar, f = 10 Hz and d = 3 cm. By using these depositiononditions, the NPs density on the substrate can be easily con-rolled varying the number of laser pulses on the Ge target (from

= 3.2 × 109 cm−2 for np = 20 (Fig. 3a) to � = 1.1 × 1011 cm−2 forp = 300 (Fig. 3d)), maintaining a fairly uniform size distributionith an average height of 〈h〉 = 2.8 ± 0.6 nm.

The evolution of the Ge NPs crystallinity was studied in detail

s a function of the background Ar gas pressure. The Raman spec-ra evolution of Ge NPs samples (f = 10 Hz, d = 3 cm and np = 300)roduced at different Pbase(Ar) values (0.5 mbar, 1 mbar, 1.5 mbarnd 2 mbar) is shown in Fig. 4. This study was performed on

r Ge NPs and Al2O3 deposition, respectively. The Ge NPs deposition on the substrateground Ar gas atoms.

as-deposited and annealed samples after a RTA treatment fordifferent annealing temperatures (Tann = 600 ◦C and 700 ◦C) andannealing times (tann = 10 s and 30 s). It is well known that the typ-ical amorphous Ge NPs Raman spectrum presents two prominenthumps located around 75 cm−1 (transverse acoustic (TA) phonons)and 280 cm−1 (transverse optical (TO) phonons), whereas a welldefined and asymmetric peak near a Raman shift of 300 cm−1

is measured for crystalline Ge NPs. In general, the NPs crys-talline quality is enhanced after performing a RTA treatment asexpected and Ge NPs crystallization onset is found for Tann = 700 ◦C,whereas no measurable crystallinity is observed for Tann = 600 ◦Cin any case, except for Pbase(Ar) = 2 mbar.that presents crystallinitybefore any thermal treatment. In particular, Ge NPs produced atPbase(Ar) = 0.5 mbar present the most amorphous character and inthis case crystallization is not achieved for tann < 30 s. However,as the Ar pressure used for the NPs production is increased, therequired annealing time for the crystallization onset is shortenedto a low value tann = 10 s. Hence, NPs crystallinity is enhanced asthe Ar background pressure is raised. The crystalline to amor-phous phase volume fraction (�c) evolution was analyzed usingthe expression �c = Ic/(Ic + Ii + ˇIa) [34,35], where Ic, Ia and Ii are theintegrated intensities corresponding to the crystalline, amorphousand intermediate (accounting for crystalline NPs size distribution)components, respectively. These parameters have been obtainedafter performing a multi-Gaussian fitting to the experimental data(Fig. 5a). The constant is the ratio of the crystalline to the amor-phous phases cross-section and it can be approximated to a valueof “1” for crystal grains with a size of a few nanometers [36].The evolution of �c as a function of the Ar gas pressure is shownin Fig. 5b for different RTA treatments. As it can be seen, thesample produced at Pbase(Ar) = 0.5 mbar presents the poorest crys-tallinity among all the samples and for all the annealing treatments,with a value �c ∼ 48% after a RTA at Tann = 700 ◦C and tann = 30 s.As the Ar pressure for NPs production is increased, a crystalline

volume fraction enhancement is observed, being larger for thehighest pressure. In particular, a value �c ∼ 60% is found for as-deposited Ge NPs produced at Pbase(Ar) = 2 mbar. and it increases upto �c ∼ 100% after a RTA for Tann = 700 ◦C and tann = 30 s. In the case

J. Martín-Sánchez et al. / Applied Surface Science 280 (2013) 632– 640 635

Fig. 2. SEM and high resolution AFM images of Ge NPs distributions varying the experimental parameters: Ar background pressure (a–d), target–substrate distance (e–g)and laser repetition rate (h–k). The average Ge NPs height is depicted as a function of Ar background pressure (l), target–substrate distance (m) and laser repetition rate (n).The color scale for AFM images is 10 nm.

636 J. Martín-Sánchez et al. / Applied Surface Science 280 (2013) 632– 640

F r diffep

oa4atacagtRotmbnspaiwm

r〈

ig. 3. AFM images of as-deposited Ge NPs on Al2O3-coated Si (1 0 0) substrates foulses (d).

f Ge NPs produced at intermediate Ar pressures (Pbase(Ar) = 1 mbarnd 1.5 mbar), the crystalline volume fraction value varies between7% and 87%, presenting a larger value than those NPs producedt Pbase(Ar) = 1.5 mbar (i.e. higher density of crystalline NPs). Fromhese results, it is found that approximately more than a half ofs-deposited Ge NPs produced for all the Ar pressures studiedan be crystallized by performing a relatively low temperaturend short-time RTA treatment (Tann = 700 ◦C and tann = 30 s), sug-esting that although as-deposited Ge NPs are mostly amorphous,hey could present some crystallinity even before performing anyTA treatment. The full-width-at-half-maximum (FWHM) valuef the experimental Raman peaks decreases as the Ge NPs crys-allinity is enhanced (Fig. 5c). Although this reduction can be

ainly attributed to NPs crystallinity enhancement (i.e. the broadand amorphous contribution is decreased), NPs coalescence can-ot be ruled out, especially for the most crystalline as-depositedamples. Moreover, after performing a RTA treatment on Ge NPsroduced for Pbase(Ar) = 2 mbar.(which already present crystallinityfter deposition), a symmetric Raman peak centered at 300 cm−1

s found that corresponds to an almost fully crystallized Ge layerhere individual NPs could have coalescenced leading to the for-

ation of big crystal grains.In Fig. 6, it is shown the C–V measurements (RT) cor-

esponding to MOS structures for Ge NPs distributions withh〉 = 2.8 ± 0.6 nm and different NPs densities (� = 1.2 × 1010 cm−2

rent laser pulses on the target: 20 pulses (a), 50 pulses (b), 150 pulses (c) and 300

and � = 1.1 × 1011 cm−2). Before contacts deposition, the samplewas subjected to a RTA treatment for Tann = 700 ◦C and tann = 30 s.The observed C–V curves are typical for MOS structures on a p-type Si substrate in the case of high frequency measurements [37].For negative voltages, the capacitance reaches its maximum value(accumulation region), while for positive voltages the capacitanceis reduced to the minimal value (inversion region). The flat-bandvoltage shift increases with the NPs density from �VFB = 0.62 V to�VFB = 1.05 V. The reference Al2O3 oxide matrix sample presents aflat-band voltage shift value �VFB ∼ 0 V as shown in the inset ofFig. 6. Hence, the appearance of a value �VFB > 0 V when Ge NPsare integrated in the Al2O3 layer can only be related with chargetrapping in the film due to the presence of NPs. The flat-band volt-age shift is accompanied by a C–V curve shift to negative voltagesin both cases, indicating that additional positive charges (holes)were collected in the state of accumulation by injection from thesubstrate, being this C–V curve shift larger as the NPs density isincreased (i.e. larger trapped holes density), which suggests a clearrelationship between charge trapping centers and NPs density. Aflat-band voltage decrease was measured by increasing the tun-nel oxide layer thickness from ttunnel = 5 nm (1.20 V and 0.32 V at

300 K and 150 K, respectively) to ttunnel = 10 nm (�VFB = 1.05 V and0.26 V at 300 K and 150 K, respectively) with a NPs density andaverage height values of � = 1.1 × 1011 cm−2 and 〈h〉 = 2.8 ± 0.6 nmrespectively, as shown in Fig. 7.

J. Martín-Sánchez et al. / Applied Surface Science 280 (2013) 632– 640 637

F RTA-(

4

imwmpmTplatsabaIoItstsef

vteg

ig. 4. Raman spectra for as-deposited (f = 10 Hz, d = 3 cm and 300 laser pulses) andb), 1.5 mbar (c) and 2 mbar (d).

. Discussion

In the PLD process, when a laser shot impinges on the Ge targetn a background gas atmosphere, the interaction of the evaporated

aterial with the atoms of the gas produces a luminescent plumehere the temperature drops quickly as it propagates due to ther-al exchange with the gas [38]. The adiabatic expansion of the

lume and the cooling of the ablated material lead to cluster for-ation by condensation in the appropriated region of temperature.

his expansion process is accompanied by a strong shock com-ression produced by the background gas atoms and the substrate

eading to an effective backward flux of material. The backward fluxnd the incoming forward flux (produced after every laser pulse onhe target) will interfere in the region between the target and theubstrate, inducing an oscillatory dynamics of particles that eventu-lly determines their size distribution. This complex dynamics wille conditioned essentially by the experimental parameters suchs background gas pressure and laser repetition rate [8,25,29,33].n addition, the final particles size distribution will also depend onther factors like target–substrate distance or set-up geometry [29].n our case, the massive and energetic big droplets resulting fromhe ablation process are stopped at the mask or follow an almosttraight forward path without any appreciable deviation (low scat-ering with background Ar atoms), whereas lightest NPs undergotrong scattering with a consequent deviation of their trajectoriesventually reaching the substrate (see Fig. 1; more details can beound in Ref. [31]).

Condensation and coalescence (growth) of particles in the

apour phase take place in high temperature regions and, ashe particles temperature drops, agglomeration processes arenhanced simultaneously with a decrease in the particles sizerowth by coalescence [30]. From our results, it is clear that the

treated Ge NPs under different Ar background gas pressures: 0.5 mbar (a), 1 mbar

interaction of the ablated material with the Ar gas atoms con-ditions not only the particles dynamics by scattering toward thesubstrate [31], but also the plasma plume temperature [39] andtherefore, cooling of the Ge particles along their path toward thesubstrate with a more favorable agglomerates formation as the NPscool down. The complex interaction between NPs in the vapourphase as a function of the experimental parameters will also con-dition their growth by coalescence and cooling. As mentionedabove, the NPs average size increases as the gas pressure is raisedfor Pbase(Ar) > 0.5 mbar (Fig. 2l) or the target–substrate distance isincreased (Fig. 2m). In both cases, this could be explained consider-ing the time needed by the cloud of NPs to reach the substrate that islonger and therefore, the probability for NPs size growth by coales-cence in the vapour phase is larger. The onset for NPs agglomeratesformation is observed for the largest pressure Pbase(Ar) = 2 mbar andtarget–substrate distance d = 5 cm (Fig. 2d and g). It can be identi-fied two regimes along the NPs path depending on the distancefrom the target: (i) coalescence dominates agglomeration up toa distance of 4 cm; (ii) agglomeration dominates coalescence fora distance ≥ 5 cm. Unexpectedly, the largest NPs average size isobtained for the lowest ambient pressure Pbase(Ar) = 0.5 mbar that isattributed to an enhanced NPs coalescence by interaction betweenthe NPs themselves with respect to background Ar atoms.

More interesting is the NPs size distribution dependence on thelaser repetition rate (Fig. 2n). Similar qualitative results have beenpreviously reported by Wang et al. [33] for Si NPs. These results areattributed to a completely different NPs interaction regime depend-ing on the laser repetition rate value for a given experimental

conditions where initially, the Ar atoms between the target and thesubstrate are in a steady-state. Once the ablation process is started,this atmosphere is perturbed and the degree of restoration of a sta-ble initial state before the following laser pulse (i.e. arrival of new

638 J. Martín-Sánchez et al. / Applied Surface Science 280 (2013) 632– 640

Fig. 5. Multi-Gaussian fitting to the experimental Raman peak for crystalline Ge NPs(a). The obtained crystalline to amorphous phase volume fraction �c as a functionof the RTA treatment for different Ar background gas pressures is shown in (b). TheFWHM Raman peaks evolution in presented in (c).

Fig. 6. C–V curves measured at RT for MOS structures with Ge NPs (〈h〉 = 2.8 ± 0.6 nm)for a NPs density value of � = 1.2 × 1010 cm−2 and � = 1.1 × 1011 cm−2 (ttunnel = 10 nm).The reference C–V curve corresponding to a reference 25-nm-thick Al2O3 layerwithout Ge NPs is shown in the inset.

Fig. 7. C–V curves measured at 150 K and RT for MOS structures with GeNPs (〈h〉 = 2.8 ± 0.6 nm and � = 1.1 × 1011 cm−2) for a tunnel layer thickness valuettunnel = 5 nm (a) and ttunnel = 10 nm (b).

Surfa

ecccrurlitlgT(roe

apitpipsgtGp

swvdtmttpicTattopt

5

wdwudasdgebt

[

[

[

[

[

[

[

[

[

[

J. Martín-Sánchez et al. / Applied

vaporated material) will determine the backward/forward parti-les fluxes interaction. Our results show a critical value fc = 10 Hzorresponding to a particles interaction regime where big parti-les formation is inhibited. In a hypothetical laser repetition valuesange where restoration is achieved (interval between two consec-tive laser pulses is larger than the time needed for initial stateestoration), the NPs size distribution should not depend on theaser repetition rate. From our experimental results and the exper-mental conditions used in this work, it is clear that this restorationime value is above 1 s (f < 1 Hz). Wang et al. found a different criticalaser repetition rate fc = 20 Hz in the case of Si NPs for an Ar back-round pressure Pbase(Ar) = 0.1 mbar and laser fluence of 4 J/cm2.his is reasonable if one considers a different particle dynamicsi.e. eventual Ge NPs distribution on the substrate) for a given laserepetition rate depending on the background gas pressure, kindf background gas, kind of evaporated material and/or amount ofvaporated material in every laser pulse (laser fluence).

The NPs crystalline quality dependence on the Ar pressure ons-deposited and annealed NPs is explained considering the plasmalume temperature dependence on the density of ablated species

nside the plume. Specifically, theoretical calculations showed thathe temperature raises as this density increases [39]. A higherlasma plume confinement (i.e. higher density of ablated species)

n a reduced volume is obtained for the highest ambient Arressures. This leads to a high interaction between the atomicpecies and/or particles resulting from the ablation of the tar-et with the Ar atoms of the ambient gas, eventually increasinghe temperature in the plasma plume. In general, it is found thate NPs crystallization is enhanced further for the largest ambientressures.

The �VFB increase as the NPs density is increased (Fig. 6) demon-trates that the charge trapping process in Ge NPs is directly relatedith the presence of the NPs in the film. Moreover, a lower �VFB

alue is found (i.e. lower charge trapping) as the temperature isecreased and the tunnel layer thickness is increased. The observedemperature dependence has been previously reported by Here-

ans et al. [40], and it is attributed to holes density available forrapping in the NPs, which is a function of the temperature andhe distance of the NPs to the substrate. Specifically, at low tem-eratures, the holes density decays exponentially as we go deeper

nto the oxide (i.e. effective holes mean free path is decreased) andonsequently, the injection of holes in the accumulation region.his leads to a holes trapping decrease in the NPs and, accordingly,

�VFB value reduction. The shorter �VFB value obtained for thehicker tunnel layer thickness is explained by the reduction of theunneling probability of holes from the substrate due to the increasef the physical Al2O3 tunnel layer thickness which reinforces ourrevious assumption about carrier storage due to the presence ofhe Ge NPs.

. Conclusions

In this work, an original shadowed off-axis deposition set-upas used to study systematically the formation of Ge NPs pro-uced by PLD in an Ar gas atmosphere at RT. The NPs distributionas analyzed as a function of the main PLD parameters withoutndesirable effects due to direct droplets/high energetic particleseposition. Uniform distributions of non-agglomerated NPs withverage height 〈h〉 = 2.8 ± 0.6 nm and a fine control in the NPs den-ity was obtained for optimized PLD conditions (Pbase(Ar) = 1 mbar;

= 3 cm; f = 10 Hz). The crystalline NPs quality has been investi-

ated by micro-Raman spectroscopy on as-deposited NPs and annhancement of their crystallinity has been demonstrated as the Arackground pressure is raised with a crystalline to amorphous frac-ion value �c > 50% for NPs produced at Pbase(Ar) = 2 mbar. The NPs

[

ce Science 280 (2013) 632– 640 639

functionality for charge trapping applications has been successfullyvalidated by electrical C–V measurements.

Acknowledgements

This work has been partially funded by: (i) FEDER fundsthrough the COMPETE program “Programa Operacional Fac-tores de Competitividade” and by Portuguese funds throughPortuguese Foundation for Science and Technology (FCT) inthe frame of the Project PTDC/FIS/70194/2006; (ii) Scientificand Technological Cooperation Program between Portugal (FCT)and Morocco (CNRST)-2010/2011; (iii) European COST MP0901-NanoTP Action. E.M.F.V. and S.R.C.P. are grateful for financialsupport through FCT/POPH grants SFRH/BD/45410/2008 andSFRH/BPD/73548/2010, respectively. J.M.-S. is grateful for financialsupport through grant CSIC JAE-DOC-070/01. The authors wouldlike to thank A. Tarazaga for scientific discussion and José Santosfor technical support.

References

[1] D. Bimberg, M. Grundmann, N.N. Ledentsov, Quantum Dots Heterostructures,John Wiley & Sons Ltd., New York, 1999.

[2] R.J. Walters, G.J. Bourianoff, H.A. Atwater, Field-effect electroluminescence insilicon nanocrystals, Nature Materials 273 (2005) 143–146.

[3] G. Conibeer, M. Green, E.C. Cho, D. König, Y.H. Cho, T. Fangsuwannarak, G.Scardera, E. Pink, Y. Huang, T. Puzzer, S. Huang, D. Song, C. Flynn, S. Park, X. Hao,D. Mansfield, Silicon quantum dot structures for tandem photovoltaic cells, ThinSolid Films 516 (2008) 6748–6756.

[4] J. Martín-Sánchez, A. Chahboun, M.J.M. Gomes, A.G. Rolo, B. Pivac, I. Capan, Car-rier storage in Ge nanoparticles produced by pulsed laser deposition, PhysicaStatus Solidi (RRL) – Rapid Research Letters 6 (2012) 223–225.

[5] S. Tiwari, F. Rana, H. Hanafi, A. Harstein, E.F. Crabbe, K. Chan, A silicon nanocrys-tals based memory, Applied Physics Letters 68 (1995) 1377–1379.

[6] Y.L. Wang, C. Chen, X.C. Ding, L.Z. Chu, Z.C. Deng, W.H. Liang, J.Z. Chen, G.S. Fu,Nucleation and growth of nanoparticles during pulsed laser deposition in anambient gas, Laser Part Beams 29 (2011) 105–111.

[7] D.B. Geohegan, A.A. Puretzky, G. Duscher, S.J. Pennycook, Time-resolved imag-ing of gas phase nanoparticle synthesis by laser ablation, Applied PhysicsLetters 72 (1998) 2987–2989.

[8] L. Patrone, D. Nelson, V.I. Safarov, M. Sentis, W. Marine, S. Giorgio, Photolu-minescence of silicon nanoclusters with reduced size dispersion produced bylaser ablation, Journal of Applied Physics 87 (2000) 3829–3837.

[9] O.M. Feroughi, C. Sternemann, Ch.J. Sahle, M.A. Schroer, J. Sternemann, H. Con-rad, A. Hohl, G.T. Seidler, J. Bradley, T.T. Fister, M. Balasubramanian, A. Sakko,K. Pirkkalainen, K. Hämäläinen, M. Tolan, Phase separation and Si nanocrystalformation in bulk SiO studied by X-ray scattering, Applied Physics Letters 96(2010) 081912.

10] Y.M. Niquet, G. Allan, C. Delerue, M. Lanoo, Quantum confinement in Genanocrystals, Applied Physics Letters 77 (2000) 1182–1184.

11] S. Codoluto, W. Baumgardner, T. Hanrath, Fundamental aspects of nucleationand growth in the solution-phase synthesis of germanium nanocrystals, Crys-tEngComm 12 (2010) 2903–2909.

12] L. Cademartiri, A. Ghadimi, G. Ozin, Nanocrystal plasma polymerization:from colloidal nanocrystals to inorganic architectures, Accounts of ChemicalResearch 41 (2008) 1820–1830.

13] M. Buljan, S.R.C. Pinto, A.G. Rolo, J. Martín-Sánchez, M.J.M. Gomes, J. Grenzer,A. Mücklich, S. Bernstorff, V. Holy, Self-assembling of Ge quantum dots in analumina matrix, Physical Review B 82 (2010) 235407.

14] H. Fukuda, S. Sakuma, T. Yamada, S. Nomura, M. Nishino, T. Higuchi, S. Ohshima,Physical and electrical properties of Ge-implanted SiO2 films, Journal of AppliedPhysics 90 (2001) 3524–3528.

15] P. Basa, G. Molnár, L. Dobos, B. Pécz, L. Tóth, A. Tóth, A. Koos, L. Dózsa, A.Nemcsics, Z. Horváth, Formation of Ge nanocrystals in SiO by electron beamevaporation, Journal of Nanoscience and Nanotechnology 8 (2008) 818–822.

16] N. Chaabane, V. Suendo, H. Vach, P.R.i Cabarrocas, Soft landing of siliconnanocrystals in plasma enhanced chemical vapor deposition, Applied PhysicsLetters 88 (2006) 203111.

17] S. Das, R.K. Singha, A. Dhar, S.K. Ray, A. Anopchenko, N. Daldosso, L. Pavessi,Electroluminescence and charge storage characteristics of quantum confinedgermanium nanocrystals, Journal of Applied Physics 110 (2011) 024310.

18] T. Kobayashi, T. Endoh, H. Fukuda, S. Nomura, A. Sakai, Y. Ueda, Ge nanocrystalsin SiO2 films, Applied Physics Letters 71 (1997) 1195–1197.

19] F. Gao, M.A. Green, G. Conibeer, E. Cho, Y. Huang, L. Pere-Wurfl, C. Flynn, Fabri-

cation of multilayered Ge nanocrystals by magnetron sputtering and annealing,Nanotechnology 19 (2008) 455611.

20] S.R.C. Pinto, M. Buljan, L. Marques, J. Martín-Sánchez, O. Conde, A. Chahboun,A.R. Ramos, N.P. Barradas, E. Alves, S. Bernstorff, J. Grenzer, A.M. Mück-lich, M.M.D. Ramos, M.J.M. Gomes, Influence of annealing conditions on the

6 Surfa

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

40 J. Martín-Sánchez et al. / Applied

formation of regular lattices of voids and Ge quantum dots in an amorphousalumina matrix, Nanotechnology 23 (2012) 405605.

21] S.R.C. Pinto, A.G. Rolo, M. Buljan, A. Chahboun, S. Bernstorff, N. Barradas, E.Alves, R. Kashtiban, U. Bangert, M.J.M. Gomes, Low-temperature fabrication oflayered self-organized Ge clusters by RF-sputtering, Nanoscale Research Letters6 (2011) 341.

22] Z. Wan, S. Huang, M.A. Green, G. Conibeer, Rapid thermal annealing and crys-tallization mechanism study of silicon nanocrystal in silicon carbide matriz,Nanoscale Research Letters 6 (2011) 129.

23] Y. Wang, X. Liao, Z. Ma, G. Yue, H. Diao, J. He, G. Kong, Y. Zhao, Z. Li, F.Yun, Solid-phase crystallization and dopant activation of amorphous siliconfilms by pulsed rapid thermal annealing, Applied Surface Science 135 (1998)205–208.

24] J. Martín-Sánchez, L. Marques, E.M.F. Vieira, Q.T. Doan, A. Marchand, A. ElHdiy, A.G. Rolo, S.R.C. Pinto, M.M.D. Ramos, A. Chahboun, M.J.M. Gomes,Ge nanocrystals with highly uniform size distribution deposited on aluminaat room temperature by pulsed laser deposition: structural, morphologicaland charge trapping properties, Journal of Nanoparticle Research 14 (2012)843.

25] T. Yoshida, S. Takeyama, Y. Yamada, K. Mutoh, Nanometer-sized silicon crystal-lites prepared by excimer laser ablation in constant pressure inert gas, AppliedPhysics Letters 68 (1996) 1772–1774.

26] G. Fu, Y. Wang, L. Chu, Y. Zhou, W. Yu, L. Han, Y. Peng, The size distributionof Si nanoparticles prepared by pulsed-laser ablation in pure He, Ar or Ne gas,Europhysics Letters 69 (2005) 758–762.

27] M.R.R. Vaziri, F. Hajiesmaeibaigi, M.H. Maleki, Monte Carlo simulation of thesubsurface growth mode during pulsed laser deposition, Journal of AppliedPhysics 110 (2011) 043304.

28] J. Barnes, N. Beer, A. Petford-Long, A. Suárez-García, R. Serna, D. Hole, M. Wey-

land, P. Midgley, Resputtering and morphological changes of Au nanoparticlesin nanocomposites as a function of the deposition conditions of the oxide cap-ping layer, Nanotechnology 16 (2005) 718–723.

29] D.H. Lowndes, C.M. Rouleau, T. Thundat, G. Duscher, E. Kenik, S. Pennycook,Silicon and zinc telluride nanoparticles synthesized by pulsed laser deposition:

[

[

ce Science 280 (2013) 632– 640

size distributions and nanoscale structure, Applied Surface Science 127–129(1998) 355–361.

30] R.C. Flagan, M.M. Lunden, Particle structure control in nanoparticle synthe-sis from the vapor phase, Materials Science and Engineering: A 204 (1995)113–124.

31] J. Martín-Sánchez, A. Chahboun, S.R.C. Pinto, A.G. Rolo, L. Marques, R. Serna,E.M.F. Vieira, M.M.D. Ramos, M.J.M. Gomes, A shadowed off-axis productionof Ge nanoparticles in Ar gas atmosphere by pulsed laser deposition, AppliedPhysics A (2012), http://dx.doi.org/10.1007/s00339-012-7131-z.

32] I. Horcas, R. Fernandez, J.M. Gómez-Rodríguez, J. Colchero, J. Gómez-Herrero,A.M. Baró, WSXM: a software for scanning probe microscopy and a tool fornanotechnology, Review of Scientific Instruments 78 (2007) 013705.

33] Y.L. Wang, W. Xu, Y. Zhou, L.Z. Chu, G.S. Chu, Influence of pulse repetition rateon the average size of silicon nanoparticles deposited by laser ablation, LaserPart Beams 25 (2007) 9–13.

34] R. Tsu, J. González-Hernández, S.S. Chao, S.C. Lee, K. Tanaka, Critical volume frac-tion of crystallinity for conductivity percolation in phosphorous-doped Si:F:Halloys, Applied Physics Letters 40 (1982) 534–535.

35] B. Zhang, S. Shrestha, S.J. Huang, P. Aliberti, M.A. Green, G. Conibeer, Structuralstudies of multilayered Ge nanocrystals embedded in SiO2 matrix fabricatedusing magnetron sputtering, Energy Procedia 2 (2010) 243–250.

36] E. Bustarret, M.A. Hachicha, M. Brunel, Experimental determination of thenanocrystalline colume fraction in silicon thin films from Raman spectroscopy,Applied Physics Letters 52 (1988) 1675–1677.

37] D.K. Schroder, Semiconductor Material and Device Characterization, John Wiley& Sons, New York, 2006.

38] G. Callies, H. Schittenhelm, P. Berger, H. Hügel, Modeling of the expansion oflaser-evaporated matter in argon, helium and nitrogen and the condensationof clusters, Applied Surface Science 127–129 (1998) 134–141.

39] M. Capitelli, F. Capitelli, A. Eletskii, Non-equilibrium and equilibrium problemsin laser-induced plasmas, Spectrochimica Acta Part B 55 (2000) 559–574.

40] P. Heremans, G. Groeseneken, H. Maes, Hole trapping during low gate bias,high drain bias hot-carrier injection in n-MOSFETs at 77 K, IEEE Transactionson Electron Devices 39 (1992) 851–857.