pulsed radiofrequency glow discharge time-of-flight mass spectrometry for nanostructured materials...

Pulsed Radiofrequency Glow DischargeTime-of-Flight Mass Spectrometry forNanostructured Materials Characterization

Marta Bustelo,† Beatriz Fernandez,† Jorge Pisonero,‡ Rosario Pereiro,† Nerea Bordel,‡

Victor Vega,‡,§ Victor M. Prida,‡ and Alfredo Sanz-Medel*,†

Department of Physical and Analytical Chemistry, Faculty of Chemistry, University of Oviedo, Julian Claverıa, 8.33006 Oviedo, Spain, Department of Physics, Faculty of Science, University of Oviedo, Calvo Sotelo.33007 Oviedo, Spain, and Scientific-Technical Services, University of Oviedo, 33006 Oviedo, Spain

Progress in the development of advanced materials stronglydepends on continued efforts to miniaturizing their struc-tures; thus, a great variety of nanostructured materialsare being developed nowadays. Metallic nanowires areamong the most attractive nanometer-sized materialsbecause of their unique properties that may lead toapplications as interconnectors in nanoelectronic, mag-netic, chemical or biological sensors, and biotechnologicallabels among others. A simple method to develop self-ordered arrays of metallic nanowires is based on the useof nanoporous anodic alumina (NAA) and self-assemblednanotubular titanium dioxide membranes as templates.The chemical characterization of nanostructured materialsis a key aspect for the synthesis optimization and thequality control of the manufacturing process. In this work,the analytical potential of pulsed radiofrequency glowdischarge with detection by time-of-flight mass spectrom-etry (pulsed rf-GD-TOFMS) is investigated for depthprofile analysis of self-assembled metallic nanostructures.Two types of nanostructured materials were successfullystudied: self-assembled NAA templates filled with arraysof single metallic nanowires of Ni as well as arrays ofmultilayered Au/FeNi/Au and Au/Ni nanowires andnanotubular titanium dioxide templates filled with Ninanowires, proving that pulsed rf-GD-TOFMS allows forfast and reliable depth profile analysis as well as for thedetection of contaminants introduced during the synthesisprocess. Moreover, ion signal ratios between elementaland molecular species (e.g., 27Al+/16O+ and 27Al+/32O2

+)were utilized to obtain valuable information about thefilling process and the presence of possible leaks inthe system.

The synthesis of structured solid-state materials with nanom-eter-sized geometries is currently receiving special interest dueto the unique physical and chemical properties exhibited by manymaterials when nanostructured (e.g., improved mechanical, opti-

cal, electrical, magnetic, or catalytic properties).1 For example,highly ordered and densely packed arrays of nanopores, nano-tubes, and nanowires2 become promising candidates for attractiveapplications in materials scientific and technological areas (suchas functionalized arrays for magnetic sensors,3 ultrahigh densitydata storage media,4 thermoelectrics,5 or spin-based electronicdevices).6 In particular, self-organized magnetic nanowires, whosediameter ranges from a few atomic distances to a few hundredsof nanometers and their lengths can vary from a few nanometersup to micrometers, are known to offer exceptional characteristicsof technological interest, due to the convergence of high spatialordering degree with the intrinsic nature of the materialssynthesized at the nanoscale.7

There are several methods to fabricate nanowire arrays,8 anda simple one is based on the use of anodized metal membranesas templates. These patterned membranes can be filled afterwardwith the desired metallic elements via electrodeposition.9 As aresult, self-ordered arrays of dense, continuous, and highlycrystalline metallic nanowires can be synthesized inside thetemplate. Nanoporous anodic alumina (NAA) nanostructures,which are formed by dense rows of highly hexagonally orderednanopores in a honeycomb like configuration, are nowadaysamong the most common templates. The reason to use these NAAtemplates as precursor lies in the relatively simple and inexpensiveprocess of self-assembling nanopore formations by aluminumanodization and the fact that well controlled patterned nanostruc-

* To whom correspondence should be addressed. Phone/fax: +34 0.985103474.E-mail: [email protected].

† Department of Physical and Analytical Chemistry, Faculty of Chemistry.‡ Department of Physics, Faculty of Science.§ Scientific-Technical Services.

(1) Yang, X.-C.; Zou, X.; Liu, Y.; Li, X.-N.; Hou, J.-W. Mater. Lett. 2010, 64,1451–1454.

(2) Feng, X.; Shankar, K.; Varghese, O. K.; Paulose, M.; Latempa, T. J.; Grimes,C. A. Nano Lett. 2008, 8, 3781–3786.

(3) McGary, P. D.; Tan, L.; Zou, J.; Stadler, B. J. H.; Downey, P. R.; Flatau,A. B. J. Appl. Phys. 2006, 99, 08B310.108B310.6.

(4) Ross, C. A. Annu. Rev. Mater. Res. 2001, 31, 203–235.(5) Trahey, L.; Becker, C. R.; Stacy, A. M. Nano Lett. 2007, 7, 2535–2539.(6) Crowley, T. A.; Daly, B.; Morris, M. A.; Erts, D.; Kazakova, O.; Boland,

J. J.; Wu, B.; Holmes, J. D. J. Mater. Chem. 2005, 15, 2408–2413. Piraux,L.; Renard, K.; Guillemet, R.; Matefi-Tempfli, S.; Matefi-Tempfli, M.; Antohe,V. A.; Fusil, S.; Bouzehouane, K.; Cros, V. Nano Lett. 2007, 7, 2563–2567.

(7) Gao, T. R.; Yin, L. F.; Tian, C. S.; Lu, M.; Sang, H.; Zhou, S. M. J. Magn.Magn. Mater. 2006, 300, 471–478.

(8) Martın, J. I.; Nogues, J.; Liu, K.; Vicent, J. L.; Schuller, I. K. J. Magn. Magn.Mater 2003, 256, 449–501.

(9) Nielsch, K.; Muller, F.; Li, A.-P.; Gosele, U. Adv. Mater. 2000, 12, 582–586.

Anal. Chem. 2011, 83, 329–337

10.1021/ac102347v 2011 American Chemical Society 329Analytical Chemistry, Vol. 83, No. 1, January 1, 2011Published on Web 12/02/2010

tures can be formed over a large surface area.10 However, NAAmembranes present some disadvantages, including insufficientchemical stability and low mechanical resistance. More recently,the synthesis of self-aligned titanium dioxide nanotube arrays hasacquired special interest due to its high yield strength at low andhigh temperatures, low density, and excellent biocompatibility.11

Additionally, the technological applications of titania nanotubearrays is increasing in photovoltaics, photoelectrolysis, photoca-talysis, and sensing, as titania is a nontoxic and noncorrosivematerial formed by a wide band gap semiconductor oxide.12,13

Fast and reliable depth chemical characterization of thenanostructured materials remains of critical importance to assistthe optimization of the synthesis procedure and to evaluate theirroutine manufacturing quality. Therefore, analytical techniquescapable of characterizing the new materials at the nanoscale arecurrently required. Surface analytical techniques with lateral anddepth resolution on the nanometer range, such as secondary ionmass spectrometry, Auger electron spectroscopy, or X-ray pho-toelectron spectroscopy, provide interesting analytical capabilitiesfor the analysis of nanostructured materials.14 However, thosetechniques require long analysis time and expensive and complexinstrumentation (need of ultrahigh vacuum), while glow discharge(GD) spectroscopy may tackle most of such problems. In fact,GDs coupled to optical emission spectrometry (OES) and massspectrometry (MS) are today widely used for direct solid analysisdue to their advantages of fast sputtering rate (micrometers/minute), high depth resolution (nanometer), multielement capabil-ity, high sample throughput, minimal matrix effects, good detec-tion limits (microgram/gram to nanogram/gram) and experimentalsimplicity with no requirement for ultrahigh vacuum.15 However,for direct solid analysis a potential drawback of GDs is their poorlateral resolution (of the order of a few millimeters).

Pulsed-GDs (PGDs), wherein each pulse generates a packetof sample atoms of the analyzed sample, are being shown to offeran attractive analytical option to the more common GD operationmode, based on continuous powering. PGDs allow different timedomains (prepeak, plateau, and afterglow) along the pulse period,related to different dominant ionization processes.16 With the useof a gated detector it is possible to select temporal intervals withinthe power pulse, in which analyte ions have higher signal-to-noiseratios. Moreover, on average, the applied power along the pulseperiod is lower compared to the continuous mode, resulting inreduced thermal stress of the solid sample to be analyzed.

GD-MS presents some significant advantages compared to GD-OES, including isotopic information and lower limits of detection.The double-focusing and quadrupole have been the most commonmass analyzers in GD-MS. However, time-of-flight (TOF) massspectrometers are increasingly utilized due to their high sampling

rate and ability to collect complete mass spectra with the sameprecision, sensitivity, and resolution regardless of the total numberof isotopes being measured.17 These characteristics make TOFMSespecially suitable to be coupled as time gated detector to PGDs.The analytical capabilities of PGDs coupled to TOFMS are beinglately studied for varied samples. For instance, GD-TOFMS havebeen successfully applied to the elemental and molecular analysisof organic compounds in gas phase,18 to the analysis of smallvolumes of molecular inorganic gases,19 and to direct analysis ofsolid materials.20,21

In this work, radiofrequency (rf) pulsed GD-TOFMS isinvestigated as a new analytical tool for the fast and accuratecharacterization of nanostructures, in particular single (Ni) andmultilayered (Au/Ni(Fe)/Au) metallic nanowires, which aregrown by templated-assisted electrochemical deposition proce-dures in self-ordered nanoporous alumina and titania nanotubessynthesized through potentiostatic anodization techniques. High-quality characterization of these specimens becomes necessaryfor a better understanding and control of their physical propertiesin order to tailor their enhanced soft magnetic properties, suchas high magnetic permeability, uniaxial anisotropy, peculiardomain pattern, and magnetoresistance, which play a key role inthe development of future magnetoelectronic devices based onwell-controlled static and dynamic domain wall displacements inthese low-dimensional systems.22-24 Therefore, depth profileanalysis of nanostructures as well as synthesis of contaminantimpurities are investigated. Moreover, the analytical potential ofthis MS technique to allow the simultaneous acquisition ofelemental and molecular information is exploited here measuringmolecular and elemental oxygen in order to discriminate betweenthe oxygen coming from the solid material (e.g., oxides) and themolecular oxygen that comes from air, contained in the nano-structure or coming from possible leaks within the instrument.

EXPERIMENTAL SECTIONSamples Preparation. Samples were produced by using a

template assisted filling method. Arrays of Ni nanowires withdifferent lengths (1.2, 2.2, and 3.8 µm) and multilayered nanowiresconsisting of alternating layers (of Au/Ni and Au/FeNi/Au) withdifferent thicknesses (from 0.8 µm up to 1.9 µm) were electrode-posited inside the nanopores of the alumina membrane. Moreover,arrays of Ni (550 nm length) were also deposited inside a titania

(10) Sulka, G. D.; Brzozka, A.; Zaraska, L.; Jaskula, M. Electrochim. Acta 2010,55, 4368–4376.

(11) Prida, V. M.; Hernandez-Velez, M.; Pirota, K. R.; Menendez, A.; Vazquez,M. Nanotechnology 2005, 16, 2696–2702.

(12) Vega, V.; Prida, V. M.; Hernandez-Velez, M.; Manova, E.; Aranda, P.; Ruiz-Hitzky, E.; Vazquez, M. Nanoscale Res. Lett. 2007, 2, 355–363.

(13) Kim, J. C.; Choi, J.; Lee, Y. B.; Hong, J. H.; Lee, J. I.; Yang, J. W.; Lee,W. I.; Hur, N. H. Chem. Commun. 2006, 5024–5026.

(14) Fernandez, B.; Costa, J. M.; Pereiro, R.; Sanz-Medel, A. Anal. Bioanal. Chem.2010, 396, 15–29.

(15) Pisonero, J.; Fernandez, B.; Pereiro, R.; Bordel, N.; Sanz-Medel, A. TrendsAnal. Chem. 2006, 25, 11–18.

(16) Harrison, W. W.; Yang, C.; Oxley, E. Anal. Chem. 2001, 73, 480A–487A.

(17) Pisonero, J.; Costa, J. M.; Pereiro, R.; Bordel, N.; Sanz-Medel, A. Anal.Bioanal. Chem. 2004, 379, 658–667.

(18) Sola-Vaquez, A.; Martın, A.; Costa, J. M.; Pereiro, R.; Sanz-Medel, A. Anal.Chem. 2009, 81, 2591–2599.

(19) Gago, C. G.; Pereiro, R.; Bordel, N.; Ramos, P. M.; Tempez, A.; Sanz-Medel,A. Anal. Chim. Acta 2009, 652, 272–277.

(20) Lobo, L.; Pisonero, J.; Bordel, N.; Pereiro, R.; Tempez, A.; Chapon, P.;Michler, J.; Hohl, M.; Sanz-Medel, A. J. Anal. At. Spectrom. 2009, 24, 1373–1381.

(21) Muniz, A. C.; Pisonero, J.; Lobo, L.; Gonzalez, C.; Bordel, N.; Pereiro, R.;Tempez, A.; Chapon, P.; Tuccitto, N.; Licciardello, A.; Sanz-Medel, A. J.Anal. At. Spectrom 2008, 23, 1239–1246.

(22) Nguyen, T. M.; Cottam, M. G.; Liu, H. Y.; Wang, Z. K.; Ng, S. C.; Kuok,M. H.; Lockwood, D. J.; Nielsch, K.; Gosele, U. Phys. Rev. B 2006, 73,140402_1–4.

(23) Pardavi-Hovath, M.; Si, P. E.; Vazquez, M.; Rosa, W. O.; Badini, G. J. Appl.Phys. 2008, 103, 07D517.

(24) Boone, C. T.; Katine, J. A.; Carey, M.; Childress, J. R.; Cheng, X.; Krivorotov,I. N. Phys. Rev. Lett. 2010, 104, 097203_1–4.

330 Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

nanotubes membrane. Table 1 collects the samples under inves-tigation. The description of the method used for the samplepreparation of filled nanoporuous alumina and titania nanotubemembranes is included as Supporting Information.

Instrumentation. Glow Discharge Time-of-Flight Mass Spec-trometer. The rf-GD-TOFMS instrument consists of a rf-GD bayunit (rf generator, matching box, rf connector, refrigerator disk,and sample mounting system with a pneumatic piston to pressthe sample against the source) from Horiba Jobin Yvon (Long-jumeau, France), coupled to a fast orthogonal time-of-flight massspectrometer (TOFWERK, Switzerland) with a microchannel platedetector (Burle Industries Inc., Lancaster, PA).25 Additionally, aninterface, consisting of two extraction cones (sampler and skimmercones), connects the GD source to the TOFMS. This interfaceallows one to extract and focus the ions as well as to reduce thepressure between the GD source and the mass analyzer. Furtherdetails of the GD-TOFMS instrument are included as SupportingInformation.

In this work, the GD was operated in the pulsed mode.Experimental conditions (650 Pa, 55 W forward power, 2 ms pulsewidth, and 4 ms pulse period) were chosen as a compromisebetween high sensitivity and good depth resolution through theanalysis of homogeneous materials with aluminum and titaniummatrixes. The module and phase of the applied rf power wereadapted to keep the reflected power to a minimum value. Analyteion signals showed their maximum intensity in the afterglowregion of the pulse. Therefore, pulsed rf-GD-TOFMS depth profilesof the nanostructured materials were always measured by inte-grating the ion signals in the afterglow region (each isotope atits maximum position).

To achieve reliable depth profiling of thin layers using GDs, itis crucial to provide conditions for stable plasma generation atthe beginning of the sputtering. This can be done throughminimizing contaminations from the sample and anode surfaces,and thus, nanostructured materials were slightly pretreated beforeGD analysis with the following procedure: (i) since the fabricationprocess of alumina and titania templates uses liquid solutions,samples were dried with hot air for 5 min to remove residual

humidity, (ii) high-purity silicon were presputtered in the GDsource for 5 min to create a Si coating on the source that couldreduce water desorption,26 and (iii) to reduce the amount of gasspecies that could be occluded in the sample and chambersurfaces, a flushing step with Ar during 3-4 min was applied tothe GD source and the sample.

Profilometer and Scanning Electron Microscopy. The depth ofthe craters produced on the samples after GD sputtering wasmeasured with a profilometer (Perth-o-meter S5P, Mahr Perthen,Germany). Two profile traces at different direction across thecenter of the crater were measured. Additionally, the length ofthe nanowires inside the alumina and titania membranes wasmeasured by scanning electron microscopy (SEM) (MEB JEOL-6100, Japan).

RESULTS AND DISCUSSIONQualitative Depth Profiles of Alumina Nanopores Filled

with Single Metal Nanowires. Figure 1 shows the qualitativedepth profiles (analyte ion intensity versus sputtering time)obtained by pulsed-rf-GD-TOFMS at the optimized experimentalconditions for three nanoporous alumina substrates synthesizedwith different nanopore lengths (1.2, 2.2, and 3.8 µm) and thenfilled with electrodeposited Ni. As can be observed, 27Al+ ionsignal always showed a fast increase at the interface betweenthe nanoporous alumina and the aluminum substrate, indicatinga good depth resolution. The 58Ni+ signal observed in the depthprofiles demonstrates the presence of this metal within thealumina nanopores. Besides, a high 58Ni+ ion signal was alwaysobserved at the beginning of the analysis when the 27Al+ ionsignal increased from zero, indicating the presence of a verythin layer of Ni on the aluminum membrane. Although thepresence of Ni was clearly observed inside the nanopores forthe three different lengths, the depth profiles showed that theNi electrodeposition process was more uniform for the samplewith 2.2 µm length since a homogeneous 58Ni+ signal wasobtained throughout the nanopore length. As an example,Figure 1b collects also the SEM image of this alumina membranefilled with Ni. The presence of small knolls of Ni is clearlyobserved on the sample surface and it could be attributed to theoverfilling of the nanopores during the Ni electrodepositionprocess, in agreement with our results obtained from pulsed-rf-GD-TOFMS profiles.

The sputtering time necessary to reach the interface betweenthe Al2O3 nanopores filled with Ni nanowires arrays and thealuminum substrate itself was proportional to the length of thenanopores. The linear relationship observed for the threeselected lengths with the sputtering time is included asSupporting Information as Figure S1a. In contrast to Ni fillednanopores, NAA with empty nanopores did not show a linearrelationship between the length and the observed sputtering time.The qualitative depth profile obtained by pulsed-rf-GD-TOFMSfor the empty NAA templates, synthesized with a nanopore lengthof 2.2 µm, is shown in Figure S1b. It can be noticed that thepresence of air in the nanopores significantly increased the

(25) Hohl, M.; Kanzari, A.; Michler, J.; Nelis, T.; Fuhrer, K.; Gonin, M. Surf.Interface Anal. 2006, 38, 292–295.

(26) Molchan, I. S.; Thompson, G. E.; Skeldon, P.; Trigoulet, N.; Chapon, P.;Tempez, A.; Malherbe, J.; Lobo-Revilla, L.; Bordel, N.; Belenguer, Ph.; Nelis,Th.; Zahri, A.; Therese, L.; Guillot, Ph.; Ganciu, M.; Michler, J.; Hohl, M.J. Anal. At. Spectrom. 2009, 24, 734–741.

Table 1. Samples Classified by Type of Membrane,Nanopore or Nanotube Length, and NanowireCompositiona

alumina template

nanopore length (µm) nanowire

1.2 Ni (filled and overflowed)2.2 Ni (filled and overflowed) and empty3.8 Ni (filled and overflowed)3.5 Au(0.8 µm)/FeNi(0.9 µm)/Au(0.9 µm)3.5 Au(0.9 µm)/Ni(1.9 µm)/Au(- -)

titania template

nanotube length (nm) nanowire

550 Ni (filled and overflowed) and empty

a The term overflowed is used to designate the nanopores ornanotubes which are totally filled up to overflowing the whole samplesurface.

331Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

sputtering time required to reach the interface compared to fillednanopores (e.g., 400 and 150 s for the 2.2 µm length, respectively)and this could be attributed to the humidity and air inside thenanopores, which produces quenching effects in the GD plasma.

As will be further discussed, the capabilities of pulsed rf-GD-TOFMS for the measurement of molecular 32O2

+ ion signals couldhelp up to identify the presence of leaks or residual air insidethe nanopores of the NAA templates.

Figure 1. Pulsed-rf-GD-TOFMS qualitative depth profiles for NAA templates filled with single Ni nanowires: (a) 1.2 µm nanopore length,(b) 2.2 µm nanopore length, (c) 3.8 µm nanopore length. Figure 1b shows also the SEM cross-sectional view of the NAA template filledwith Ni.


Qualitative Depth Profiles of Alumina Nanopores Filledwith Multilayered Metal Nanowires. Concerning the analysisof multilayered metallic nanowires, two samples were investigated.Figure 2a shows the qualitative depth profile obtained by pulsed-rf-GD-TOFMS for the NAA template of 3.5 µm nanopore lengthfilled with three electrodeposited metal layers of Au (0.8 µm), FeNialloy (0.9 µm), and Au (0.9 µm). As can be seen, the layersdeposited inside the nanopore can be discriminated with a gooddepth resolution. In particular, the first deposited Au (inner layer)had a narrower distribution compared to the outer Au layer, while56Fe+ and 58Ni+ ion signals showed a perfect match along thedepth profile. As reported above for the Ni filled nanopores, ahigh 197Au+ ion signal was observed at the beginning of theanalysis, followed in this case by a drop in the ion signal,indicating a nonhomogeneous distribution of Au through theexternal film of the nanopore. Figure 2b shows the SEM imageof this nanostructured material being in agreement with theexperimental results obtained from pulsed rf-GD-TOFMS profiles.Thus, GD and SEM measurements were allowed to obtaincomplementary information, confirming that the distribution ofAu in the external layer of the nanopores was not homogeneousand, also, that the thickness of the two Au layers was not the

same. Additionally, it should be highlighted that the analysis timerequired to sputter the nanowires reaching the aluminum sub-strate was only about 4 min, so pulsed rf-GD-TOFMS can beemployed as a powerful tool for the fast and reliable characteriza-tion of nanostructured materials.

The qualitative depth profile obtained by pulsed-rf-GD-TOFMSfor the same NAA template (3.5 µm nanopore length) filled witha different configuration of metallic electrodeposited layers (0.9µm Au, 1.9 µm Ni and Au up to covering the sample surface) canbe observed in Figure 3a. In this case, the sputtering time requiredto reach the aluminum substrate was also about 4 min. However,although the intermediate Ni layer and the internal Au film werediscriminated with a good depth resolution, the external elec-trodeposited Au layer could not be appropriately distinguished: ahigh 197Au+ ion signal was observed at the beginning of theanalysis, indicating a very thin film of Au at the sample surface,but the Au layer (which during the synthesis process wasthought to be of 0.9 µm length) was not found in this qualitativeprofile. This fact was validated by SEM measurements. Figure3b shows the SEM image obtained and, as can be seen, only twoclear Au and Ni layers were deposited inside the aluminananopores.

Figure 2. Depth profile characterization for the NAA template of 3.5 µm nanopore length filled with trilayer nanowires (Au/FeNi/Au) made ofAu (0.8 µm), FeNi (0.9 µm), and Au (0.9 µm): (a) qualitative in-depth profile obtained by pulsed rf-GD-TOFMS and (b) SEM cross-sectional viewusing backscattered electrons to enhance the compositional contrast.


The failure in the expected electrodeposition of the externalAu layer could be related to the different metal depositionprocesses within the nanoporous membrane. As can be observedin Figure 3b, with the work with electroplated Au/Ni layers,loosening problems and cracked membrane pieces were observedafter the Ni deposition. The enlargement of sample surface imagein Figure 3b shows several zones where the membrane wasbroken and the aluminum substrate appeared at the samplesurface. In such a way, during the deposition of the second andexternal Au layer, the electrolyte could be preferably depositedin zones where the alumina membrane has been removed,because the electric field will be locally more intense at thealuminum substrate than at those zones where the NAA mem-brane was still present. In other words, it seems clear that a fastand reliable depth characterization of nanopores quality, filled withsingle or multilayered metals, is of critical importance to assistthe optimization of the electrodeposition procedures, as well as

to evaluate their routine manufacturing quality. Of course, resultsshown here provide evidence that pulsed rf-GD-TOFMS consti-tutes a promising technique for fast quality assurance tests.

Qualitative Depth Profiles of Titania Nanotubes Filledwith Single Metal Nanowires. Alternatively, titania nanotubemembranes filled with single Ni nanowires in the nanometer rangewere also investigated. Figure 4 shows the qualitative depth profileobtained by pulsed rf-GD-TOFMS for a titania nanotube of 550nm nanopore length filled with Ni. As can be observed, 58Ni+ and48Ti+ ion signals showed a sharp interface, indicating a gooddepth resolution between the nanotubes and the titaniumsubstrate. Moreover, the Gaussian distribution observed forthe 58Ni+ profile suggests a good electrodeposition process anda homogeneous distribution of Ni inside the nanotube. Al-though not shown here, SEM images obtained for this samplealso showed that the nanostructured membrane was full andoverflowed with Ni. Interestingly, the time necessary to sputter

Figure 3. Depth profile characterization of the NAA template of 3.5 µm nanopore length filled with trilayer nanowire arrays of Au (0.9 µm), Ni(1.9 µm), and Au: (a) qualitative in-depth profile obtained by pulsed rf-GD-TOFMS and (b) SEM cross-sectional views using backscatteredelectrons to enhance the compositional contrast.


the titania nanotubes filled with Ni nanowires array (indicatedby the remarkable increase of 48Ti+ signal and the decrease ofthe 58Ni+ signal) was only about 15 s. Again, the sputteringtime required to reach the sample substrate for the nonfillednanopores increased compared to the Ni filled. However, inthis case the sputtering time required to reach the Ti substratewas just 20 s for the empty nanotube. In other words, with Tiinstead of Al the sputtering rates were not so different for theNi-filled and empty titania membranes. This could be ascribedto the shorter length of the nanotubes (nanometer) and,therefore, to the comparatively lower air presence in compari-son with the micrometric sized nanopores in Al substrates(NAA templates).

Analysis of Metal Impurities by Pulsed rf-GD-TOFMS.One of the recognized advantages that GD-TOFMS could offercompared to GDs coupled to OES are lower limits of detectionfor some specific elements as well as its ability to simultaneouslycollect complete mass spectra. On the other hand, analysis ofpossible impurities in nanostructures is a critical aspect for theoptimization of the sample preparation and electrodepositionprocesses. Therefore, quality control of impurities is mandatoryfor their subsequent applications. Figure 5 shows the mass spectraobtained at the optimized pulsed rf-GD conditions for twonanostructured samples based on alumina template with 2.2 µmlong nanopores (filled with Ni nanowires and empty) at twodifferent positions along the GD-TOFMS profile: inside thealuminum substrate (Figure 5a) and at the sample surface (Figure5b,c). As can be observed in Figure 5a, at the selected massinterval, the ion signals from 206Pb+, 207Pb+, and 208Pb+ wereclearly observed, indicating that the aluminum used for thepreparation of NAA membranes contained also Pb (the nominalcomposition of Pb in the Al foil was about 25 µg/g).

On the other hand, significant differences can be found in themass spectra at the sample surface region for the empty and Nifilled nanopores (parts b and c of Figure 5, respectively). As canbe seen in Figure 5b, 206Pb+, 207Pb+, and 208Pb+ signals togetherwith a small 197Au+ signal were observed at the sample surfacefor the empty nanopores, which could be attributed to thepresence of Pb impurities in the aluminum substrate and thepossible contamination of the Pt electrode with Au at low trace

concentration levels (i.e., during the fabrication step of NAAmembranes by anodization using the Pt electrode, Au was alsodeposited on the sample surface). Nevertheless, for the Ni fillednanopores, not only 206Pb+, 207Pb+, and 208Pb+ ion signals whereobserved but also ion signals from 194Pt+, 195Pt+, 196Pt+, 197Au+,198Hg+,199Hg+, 200Hg+, 201Hg+, 202Hg+, and 204Hg+ were clearlyidentified in the mass spectrum (see Figure 5c). In this case,the presence of Pt, Au, and Hg at the sample surface and not inthe aluminum substrate could be attributed not only to anodizationbut also to electrodeposition procedures employed in the samplepreparation as well as to trace impurities on the electrodes andliquid solutions used. Therefore, pulsed rf-GD-TOFMS provesagain to be a powerful tool for the depth characterization ofnanostructured materials, offering information of the possiblecontamination sources (even at very low concentration levels) thatcould affect the final properties and quality of fabricated nano-structured devices. Different types of impurities were identifiedat the aluminum substrate and at the Ni filled nanowires.Therefore, two possible contamination sources were identified tobe checked for final quality assurance: the purity of the samplesubstrate and reagents and process of the membrane synthesis,respectively.

Small Molecules Information Potential: Analysis of Atomicand Molecular Oxygen Ions. Besides the good sensitivityachieved by pulsed rf-GD-TOFMS for the analysis of elementalimpurities at trace concentration levels, the use of pulsed rf-GDscombined to a TOF mass spectrometer with time-gated detectionoffers the possibility of obtaining both elemental and molecularinformation using different pulse regions (i.e., prepeak, plateau,and afterpeak).18 A preliminary study was then carried out here,aiming to identify possible sources of 16O+ and 32O2

+ ion signals:the alumina membrane (i.e., Al2O3) and the air (molecularoxygen) existing inside the nanopores or as a leak of theexperimental system. Although several limitations were foundto identify the origin of measured 16O+ and 32O2

+ signals, dueto fragmentation as well as recombination processes that maytake place at the different pulse regions (especially at theafterglow where 16O+ from alumina could be recombined to32O2

+, while the 32O2+ from air could be partially fragmented

as16O+), promising preliminary results obtained are reportedhere. Figure 6 shows the observed 27Al+/16O+ and 27Al+/32O2

+

ratios (logarithmic scale) obtained along the pulse (2 ms) atan intermediate depth for a Ni filled NAA membrane and foran empty NAA membrane of 2.2 µm length. As can be observedin the figure, the 27Al+/16O+ and 27Al+/32O2

+ ratios in the plateauwere similar for the Ni filled template, whereas the 27Al+/16O+

ratio was significantly higher than the 27Al+/32O2+ ratio for the

empty nanopores thus indicating a higher 32O2+ content (as

expected) in this latter case. Therefore, the efficiency of theNi electrodeposition processes could be investigated throughthe 27Al+/16O+ and 27Al+/32O2

+ ratios since the lower the 27Al+/32O2

+ ratio, the higher the presence of air (e.g., due to anonhomogeneous metals distribution inside the nanopore orto the failure in the electrodeposition process, as was previouslyobserved in Figure 3a for the multilayer nanostructure).

CONCLUSIONSThe interesting capabilities of pulsed-rf-GD-TOFMS for the

chemical depth profile characterization of highly ordered and self-

Figure 4. Qualitative in-depth profile obtained by pulsed rf-GD-TOFMS for a titania membrane filled with single Ni nanowires of 550nm length.


assembled nanostructured oxide templates with metallic nano-wires deposited inside were here successfully demonstrated.Advantages of this fast direct solid analysis technique (not yet in

the market) for its use in the analytical characterization of thosenanostructured samples include simple sample preparation stagesbefore the measurement, good depth profile resolution, and high

Figure 5. Mass spectra between m/z 190 and 215 of an alumina membrane with 2.2 µm nanopore length. Discharge conditions: 650 Pa, 55W, 250 Hz pulse frequency and 2 ms pulse width: (a) mass spectrum for a Ni filled NAA template at the aluminum substrate; (b) mass spectrumfor an empty NAA template at the sample surface position; (c) mass spectrum for a Ni filled NAA template at the sample surface position.


sensitivity to reveal the presence of contaminants during thesynthesis process. Besides, ion signal ratios between elementaland molecular species allowed one to obtain valuable informationabout the homogeneity of the filling process and the presence ofpossible leaks in the system.

Therefore, this new analytical tool offers a tremendous interestto assist the synthesis optimization process as well as for thequality control of nanowires growth in patterned nanostructuredmembranes. This work warrants further research to uncoverfurther potentialities of the pulsed-rf-GD-TOFMS for its applicationin nanotechnology or in the “nanoworld”.

ACKNOWLEDGMENTFinancial support from Spanish Ministry of Science and

Innovation and FEDER Programme through Grant MAT2007-

65097-C02 as well as from Consejerıa de Educacion y Ciencia delPrincipado de Asturias (Ref COF08-10 and Ref FC09-IB09-131) isacknowledged. B. Fernandez and J. Pisonero acknowledge finan-cial support from the “Juan de la Cierva” and “Ramon y Cajal”Programs of the Ministry of Science and Innovation of Spain,respectively. Finally, we especially thank the contract with HoribaJobin Yvon for the loan of the GD-TOFMS instrument.

SUPPORTING INFORMATION AVAILABLEAdditional information as noted in text. This material is

available free of charge via the Internet at http://pubs.acs.org.

Received for review September 14, 2010. AcceptedNovember 18, 2010.

AC102347V

Figure 6. 27Al+/16O+ and 27Al+/32O2+ ratios obtained along the pulse profile at two different positions of the qualitative in-depth profile at the

center of the nanopores, both for Ni filled and empty nanopores of a NAA template with 2.2 µm pores length.


pulsed radiofrequency glow discharge time-of-flight mass spectrometry for nanostructured materials...

Documents