Wafer After Contact Electrification Using Kelvin Force Microscopy

Manila Journal of Science 10 (2017), pp. 114-125

Surface Analysis of Silicon (100) Wafer After Contact Electrification Using Kelvin Force Microscopy

Jose M. Esmeria, Jr.,1, 2 Romeric Pobre1

1Physics Department, De La Salle University, 2401 Taft Avenue, Manila 1004, Philippines 2 Reliability Engineering Department, TDK Philippines Corporation (TPC), 119 East Science Avenue, SEPZ, Biñan, Laguna 4024, Philippines.

*Corresponding Author: [email protected]

ABSTRACT

Contact electrification was demonstrated on Si (100) wafer, and surface charge images at submicron scale were analysed using Kelvin force microscopy (KFM). Potential map images have shown carpet-like patterns on the (100) plane of Si wafer. Individual potential spikes that appeared on the surface are indicative of the presence of charges arising from contact electrification. It was clearly shown that positive and negative surface potential maps on Si (100) wafer with low resistivity have minimal change in the order of ±2.5 mV after 4800 seconds of noncontact electrification. The mechanism for the slow discharged on the Si (100) wafer can be modelled like a clamped capacitor direct current (D.C.) electric circuit.

Keywords: Surface interfaces, Contact electrification, Triboelectricity.

INTRODUCTION force, room humidity, and ground continuity of the material (Hogue, 2004). A significant number of researches have Triboelectric charging between two been done for the last 50 years in contact different metals is widely accepted as an electrification (CE). CE or triboelectricity has exchange of electrons due to their work function beneficial applications in medical sensors difference. This difference is brought about by , body implants, pharmaceuticals and the difference of the individual Fermi levels of photocopying (Lacks & Mohan Sankaran, both metals where some electrons flow from 2011), and microelectric generators (Zhou metal with higher Fermi level into the other et al., 2013). However, charges generated metal (Williams, 2012). It was also reported through CE are a prelude to a charge device that triboelectric charging for polymers model (CDM) type of electrostatic discharge can be best explained by two well-known (ESD) damage to electronic devices. Typical mechanisms, one involving electron transfer ESD damage associated with CDM is <125 V and the other due to the presence of ions on for class C1 devices (“Part 5: Device Sensitivity the surface of the material. Electron transfer and Testing » Electro Over-Stress (EOS) /ESD on an insulator was reported to be a function of Association, Inc.,” n.d.). Moreover, HDD read/ its physical condition, and surface impurities, write sensors have even lower threshold at <1 such as ion contaminants, could well influence V (Baril, Nichols, & Wallash, 2002). CE is a tribocharging (Diaz & Felix-Navarro, 2004). well-known phenomenon, but the fundamental The use of atomic force microscopy has mechanisms behind CE are not yet fully made advances in the in the study of contact understood (Lacks & Mohan Sankaran, electrification at nanometer levels (Gady, 2011). Current understanding of CE revolves Reifenberger, & Rimai, 1998). By attaching around the idea that when two materials are 5-µm spheres on the AFM tip, these spheres brought into surface contact and made to act as CE applicators on dielectric substrates. separate, each surface will produce uniform With this early setup, contact electrification but oppositely electric charge distribution. was studied using force interactions between Depending on the type of material, electric the sphere and the sample. Later, potential charges may decay rapidly upon contact to mapping (Melitz, Shen, Kummel, & Lee, ground (or ionization) for dissipative materials 2011) by Kelvin force microscopy or KFM—a or charges may stay longer for insulators. A special function of atomic force microscope or common way to determine material’s charge AFM—were mostly utilized to characterize affinity—either positive or negative—is surface electric potentials. KFM was also thru the use of triboelectric series (Diaz & instrumental in demonstrating mosaic Felix-Navarro, 2004). Triboelectric series is patterns of both positive and negative charge based from empirical method that sequences for each material after contact for polymer materials based from their charge transfer specimens (Baytekin, Patashinski, Branicki, & from the positive (+top) to the negative Baytekin, 2011). Aside from the mosaic charge (−bottom). When two different materials patterns, of both positive and negative charge contact and separates (CE), the material distributions, the study also showed material listed at positive (+) side will likely be charged transfer and changes in surface composition positively and the negative (−) side will be occurred after contact electrification. Another charged negatively. However, the amount of study suggested that these mosaic patterns electric charge generated on the surface of two are thought to be an outcome of some complex materials after CE is dependent on the applied mechano-chemical reactions (Sakaguchi, ( ) Therefore, it is in the interest of this work ( ) = [ + + ] to study the effects of contact electrification 1 2 𝜕𝜕𝜕𝜕 𝑧𝑧 ( ) of aTherefore, semicond uctorit is in surface the interest using of a thiscommon work 𝐹𝐹𝑒𝑒𝑒𝑒(𝑧𝑧) = − 2[ 𝑉𝑉𝑐𝑐𝑐𝑐+ 𝑉𝑉𝑑𝑑+𝑑𝑑 𝑉𝑉𝑎𝑎𝑎𝑎𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠] 𝑠𝑠 𝜕𝜕𝜕𝜕 (4.0) touchto -studyup tool the and effects propose of contact a model electrification for the 1 2 𝜕𝜕𝜕𝜕 𝑧𝑧 Therefore, it is in the interest of this work 𝑒𝑒𝑒𝑒 𝑐𝑐𝑐𝑐 𝑑𝑑𝑑𝑑 ( 𝑎𝑎𝑎𝑎) chargeof a semiconddecay atuctor the surfacesurface using of Si a commonsample( ) = 𝐹𝐹[ 𝑧𝑧+ −+2 𝑉𝑉 ]𝑉𝑉 𝑉𝑉 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝑠𝑠 𝜕𝜕𝜕𝜕 to study the effects of contact electrification 1(4.0) 𝜕𝜕𝜕𝜕 𝑧𝑧 touch-up tool and propose a model for the 2 of aafter semicond CE.uctor surface using a common 𝐹𝐹𝑒𝑒𝑒𝑒 𝑧𝑧 − 2 𝑉𝑉𝑐𝑐𝑐𝑐 𝑉𝑉𝑑𝑑𝑑𝑑 𝑉𝑉𝑎𝑎𝑎𝑎𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝜕𝜕 𝜕𝜕 charge decay at the surface of Si (4.0)sample Thus, potential map images have two touch -up tool and propose a model for the after CE. components: tip bias ( and ) and 116 MANILA JOURNAL OF SCIENCE charge decay at the surface VOLUMEof Si sample 10 (2017) Thus, potential map images have two KFM after CE. capacitance gradient ( ( )) . Thus, Thus, potentialcomponents: map images tip bias have ( two𝑎𝑎𝑎𝑎 and 𝑑𝑑) 𝑑𝑑 and KFM capacitance gradient is 𝑉𝑉influenced𝑉𝑉 by the Makino, Ohura, & Iwata, 2014). These ions KFM Surface potential microscopy or components:SPM capacitance tip bias ( gradient and )( and( ) ) . Thus, KFM surface charge density on𝑉𝑉𝑎𝑎𝑎𝑎𝜕𝜕𝜕𝜕 the𝑧𝑧 surface⁄𝑉𝑉𝑑𝑑𝜕𝜕𝑑𝑑𝜕𝜕 of the on the surface are formed from molecular bond (Melitz, Shen, Kummel, & Lee, 2011)capacitance is a capacitancegradient ( gradient( )) .is Thus,influenced by the Surface potential microscopy or SPM 𝑎𝑎𝑎𝑎 𝑑𝑑𝑑𝑑 capacitancesample. surfacegradient charge Therefore,is 𝑉𝑉influenced density 𝑉𝑉the byon𝜕𝜕𝜕𝜕 thethechange𝑧𝑧 surface⁄ 𝜕𝜕 𝜕𝜕in ofsurface the breaking that arises after triboelectrification SurfacespecialSurface (potentialMelitz, functionpotential Shen, microscopy mofKummel,icroscopy an AFM,or SPM &or Leeand SPM(Melitz,, it 2011) is alsois a surface chargechargesample. density densityTherefore, on𝜕𝜕𝜕𝜕 the𝑧𝑧 changessurface⁄ 𝜕𝜕the𝜕𝜕 of changethe the force in onsurface the tip where radicals are formed on each opposing Shen,(Melitz,known Kummel,special Shen, commonly Kummel,function & Lee, as& of 2011) LeeKFMan , AFM, 2011)is, whicha specialisand a produceit is alsos sample. Therefore,charge( ) and densitythe thchangeerefore changes in surfacechanges the force the on thepotential tip functionspecial function of an AFM,of an AFM, and itand is alsoit is also known surfaces (Mazur & Grzybowski, 2017, p. 2025). imagesknown of potentialcommonly mapsas KFM from, which the scannedproducecharge s density (changes) the force on the tip known commonly as KFM, which produces image andof the th surface.erefore changes the potential Studies on contact electrification using EFM commonlysurfaceimages as. KFM,The of potential whichcantilever produces maps tipfrom images function the ofscanned ( )is and 𝐹𝐹th𝑒𝑒𝑒𝑒erefore𝑧𝑧 changes the potential Commented [LJA2]: Please check if changes reflect the images of potential maps from the scanned image of the surface. intended meaning. OK or KFM were mostly related to insulators and potentialrelatedsurface maps to .from theThe energythe cantilever scanned of the surface. tipcapacitance function Theimage Cis of the𝐹𝐹 surface.𝑒𝑒𝑒𝑒 𝑧𝑧 Commented [LJA2]: Please check if changes reflect the surface. The cantilever tip function is 𝑒𝑒𝑒𝑒 AFM commonly uses a tappingCommented mode to[LJA2]: Pleaseintended check meaning. if changes OK reflect the prepared polymers, which show potential cantileverrelated tip function to the isenergy related of to the the capacitanceenergy𝐹𝐹 𝑧𝑧 C intended meaning. OK relatedbetween to the theenergy tip of theand capacitance the sample C . The AFM commonly uses a tapping mode to between the tip and the sampleAFM. The commonly scan usesthe asurface tapping (mode“Basic to Theory Atomic maps of the surfaces using the tip as a contact ofbetween theelectrostatic capacitance the tip Cfandorce between the (sample the) between tip. andThe thethe tip scan the surface (“Basic Theory Atomic electrostatic force ( ) between thescan tipthe surfaceForce Micros(“Basiccopy Theory (AFM Atomic),” n.d.) to obtain its electrifying tool. These KFM studies used in sample.electrostaticand Thesample electrostaticforce surface ( ) between forceis then between the related tip the to the Force Microscopy (AFM),” n.d.) to obtain its and sample surface𝑒𝑒𝑒𝑒 is then related Forceto the Micros physicalcopy (AFM topography),” n.d.) to obtain. In this its method, the tip situ techniques to measure triboelectrification tipand andrate sample sample of changes surface surface isof thenenergyis𝐹𝐹 then related𝑒𝑒𝑒𝑒𝑧𝑧 thatrelated to theto the physical topography. In this method, the tip rate of changes𝐹𝐹𝑒𝑒𝑒𝑒 𝑧𝑧 of energy𝐹𝐹 𝑧𝑧 that physical topographylightly .touches In this method, the surfaces the tip (thus tapping) after frictional-force applications by utilizing raterate of of changeschanges ofof energyenergy that that lightly touches the surfaces (thus tapping) lightly touches the surfaces (thus tapping) an AFM cantilever tip. These techniques were ( ) at its oscillating frequency (60 to 100 kHz) ( ) =( ) (() ) at its oscillatingat its frequency oscillating (60 frequency to 100 kHz) (60 to 100 kHz) ( ) = ( ) =( ) (1.0)( ) (1.0) (1.0) while maintaining constant oscillation able to perform a comparative analysis of 1 𝑑𝑑𝑑𝑑 𝑧𝑧 (1.0) while maintainingwhile maintainingconstant oscillationconstant oscillation 1 𝑑𝑑𝑑𝑑 𝑧𝑧 12𝑑𝑑𝑑𝑑 𝑧𝑧 2 2 triboelectrification and contact electrification 𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒 amplitude amplitudeamplitudeto generate totothe generate generateimage. The thethe image. image. The The 𝐹𝐹 𝑧𝑧𝐹𝐹 𝐹𝐹−𝑧𝑧𝑒𝑒𝑒𝑒2 𝑧𝑧𝑑𝑑𝑑𝑑− 2∆−𝑉𝑉2𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑 ∆𝑉𝑉∆𝑉𝑉 (Zhou, Li, Niu, & Wang, 2016) for a dielectric wherewherewhere where isis ththe eis voltage voltagethis eth voltagee differencevoltage difference difference difference between between between betweent apping modettappingapping produces modemode high producesproduces-resolution high high-resolution-resolution ( ( )) sample. AFM-KFM was also used to analyze tiptip andtipand tip and sample.sample. and sample. sample. Where Where Where is the ( ( capacitance( (is)) )the) imaging isis thethe withoutimagingimaging leaving withoutwithout artefact leaving leaving or damage artefact artefact or ordamage damage capacitance∆𝑉𝑉 ∆ 𝑉𝑉gradient∆𝑉𝑉 with respect to z, on the surface of the samples. The semiconductor materials like SiO under gradientcapacitancecapacitance with respect gradient togradient z, which with withis the respect verticalrespect toto zz,, onon the surfacesurface ofof thethe samples.samples. The The 2 which is the vertical 𝜕𝜕𝜕𝜕distance𝑧𝑧 ⁄ 𝜕𝜕𝜕𝜕between whichwhich is isthe the vertical vertical 𝜕𝜕𝜕𝜕distance 𝜕𝜕𝜕𝜕distance𝑧𝑧 𝑧𝑧 ⁄⁄ 𝜕𝜕 𝜕𝜕between𝜕𝜕betweentopographic topographictopographic image generated imageimage by generated generatedtapping by by tapping tapping atmospheric conditions (Zhou et al., 2013) as distancesample and between tip. Whereas sample and is definedtip. Whereas as ismode AFM is often called height image. sample and tip. Whereas is defined as mode AFM is often called height image. well as n-type GaAs (Brunkov et al., 2013), defined sample as and tip. Whereas is defined as mode AFM is often called height image. which had a measured surface potential of 6 = ( ) + + ∆𝑉𝑉 (2.0) SPM on the other hand utilizes the same = ( ) + + ∆𝑉𝑉 = ( ) + + ∆𝑉𝑉 (2.0) (2.0)tapping initiallySPMSPM toon generate thethe otherother height hand hand images. utilizes utilizes the the same same mV. Furthermore, similar material (Shiota, 𝑎𝑎𝑎𝑎 𝑑𝑑𝑑𝑑 𝑐𝑐𝑐𝑐 (2.0) ∆𝑉𝑉 𝑉𝑉 𝑡𝑡 𝑉𝑉 ( ) 𝑉𝑉 tapping initially to generate height images. n.d.) using Si (111) was performed at ultrahigh where and 𝑎𝑎𝑎𝑎 are 𝑑𝑑the𝑑𝑑 DC𝑐𝑐𝑐𝑐 bias Height imagestapping are stored initially, and tothe generate tip then height images. voltage wonhere∆ 𝑉𝑉the∆ 𝑉𝑉tip 𝑉𝑉𝑎𝑎𝑎𝑎, 𝑉𝑉𝑡𝑡 and( 𝑡𝑡) = 𝑉𝑉𝑑𝑑𝑑𝑑𝑉𝑉 ( )𝑉𝑉 𝑐𝑐𝑐𝑐is 𝑉𝑉are the AFMthe DClifts bias up to Heighta certain images height are point stored called, and z the tip then vacuum (UHV) AFM by scratching the surface wherewhere𝑑𝑑 𝑑𝑑 and are and𝑎𝑎𝑎𝑎 the DC( ) bias are voltagethe DC on bias Height images are stored, and the tip then oscillatingvoltage𝑉𝑉 drive on voltage𝑉𝑉 the𝑡𝑡 tip on, tip( and) = is the is theheight. AFM Utilizing lifts upthe tostored a certain image fromheight the point called z of Si the with an uncoated AFM Si tip as a CE the tip,voltage is the on AFMthe𝑉𝑉𝑑𝑑𝑑𝑑𝑉𝑉𝑎𝑎𝑎𝑎 tip oscillating𝑡𝑡 , 𝑉𝑉𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑎𝑎𝑎𝑎( )𝑡𝑡𝑠𝑠= drive voltage is the AFM lifts up to a certain height point called z resonantoscillating frequency𝑉𝑉𝑑𝑑𝑑𝑑 drive of 𝑉𝑉the𝑎𝑎𝑎𝑎 voltage 𝑡𝑡cantilever on tip, and and previous is the tappiheight.ng modeUtilizing scan, the the stored entire image from the generator, in which a negative charge pattern on tip oscillatingis contactand is potential the drive resonant voltagedue to𝑉𝑉𝑎𝑎𝑎𝑎 frequencywork 𝑡𝑡on function𝜔𝜔tip𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 and of𝑠𝑠 the surface is the is height.scanned Utilizing again atthe astored fixed image from the resonant frequency𝑉𝑉𝑎𝑎𝑎𝑎 𝑡𝑡of the𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 cantilever𝑠𝑠 , and previous tapping mode scan, the entire was observed and measured to be −0.1 V at cantilever,differenceresonant betweenand frequency is tipcontact ( )of and potential the sample cantilever due to𝜔𝜔amplitude , and setprevious point A Otappi, whichng followed mode thescan, the entire 𝑐𝑐𝑐𝑐 is contact potential due to work function surface is scanned again at a fixed and𝑉𝑉 is expressed as 𝜔𝜔potential gradient of the surface. This its peak. work function idifferences contact difference betweenpotential𝑡𝑡 betweentip due ( to) andwork tip 𝑠𝑠 (andsample function surfaceamplitude is set scanpointned AO , whichagain followedat a thefixed 𝑐𝑐𝑐𝑐 ∅ ∅ 𝑉𝑉 ) second image is then interleaved withO the Si is a common semiconductor material and sampledifference andand( isis expressedexpressedbetween) asastip ( and sample amplitudepotential gradientset point of A the, which surface. followed This the 𝑐𝑐𝑐𝑐 𝑡𝑡 first image𝑠𝑠 to generate surface potential is widely used as a base material for almost and𝑉𝑉 is= expressed as (3.0)∅ ∅ potentsecondial image gradient is then ofinterleaved the surface. with theThis ∅𝑠𝑠−∅𝑡𝑡 ( ) ∅𝑡𝑡 image∅𝑠𝑠 as shown in Figure 1. all electronic devices. These novel techniques 𝑐𝑐𝑐𝑐 𝑞𝑞 = (3.0) (3.0) secondfirst image image to isgenerate then interleaved surface potential with the 𝑉𝑉 ( ) q is the electronic charge.∅𝑠𝑠−∅𝑡𝑡 Substituting (3.0) using the tip for contact electrification can = (3.0) Figufirstimagere 1. Conceptimage as shown ofto AFM ingenerate F-igureKFM. 1 . surface Commentedpotential [LJA3]: Insert Figure 1 here. to (2.0) and then𝑐𝑐𝑐𝑐 to (1.0), therefore, the quantify both the charge distribution and the q is the electronic𝑉𝑉 ∅𝑠𝑠−∅𝑡𝑡𝑞𝑞 charge. Substituting image as shown in Figure 1. electrostaticq is the force electronic ( ) exerted charge. between Substituting (3.0) 𝑉𝑉𝑐𝑐𝑐𝑐 𝑞𝑞 The SPM (KFM)Fig uimagingre 1. Concept techniques of AFM -KFM. Commented [LJA3]: Insert Figure 1 here. force applied (Cai & Yao, 2016); however, a (3.0)the tipto and(2.0)to (2.0) sample and and then can then beto torepresented(1.0), (1.0), therefore, therefore, as the the q is the electronic𝑒𝑒𝑒𝑒 charge. Substitutingused (3.0) in this experiment will give some real-world scenario depends on the common electrostatic(Melitz et.al.electrostatic, 2011) force 𝐹𝐹 exertedforce𝑧𝑧 between ( ) exerted the between tip Figure 1. Concept of AFM-KFM. Commented [LJA3]: Insert Figure 1 here. to (2.0) and then to (1.0), therefore, theinsights into changesThe SPM of (KFM)the material imaging techniques material in contact. An industrial-type Q tip and samplethe cantip and be represented sample can as be ( Melitzrepresented et.al., as electrostatic force 𝑒𝑒𝑒𝑒( ) exerted betweensurfaces afterused contact in thiselectrification, experiment using will give some or cotton tip, for example, is a common material 2011) (Melitz et.al., 2011) 𝐹𝐹 𝑧𝑧 The SPM (KFM) imaging techniques the tip and sample can be represented as insights into changes of the material for touch-up and removal of contamination 𝐹𝐹𝑒𝑒𝑒𝑒 𝑧𝑧 3 used in this experiment will give some (Melitz et.al., 2011) ( ) surfaces after contact electrification, using in electronicsTherefore, manufacturing. it is in the Theseinterest oftools this work ( ) = [ + + ] (4.0) insights into changes of the material to study the effects of contact electrification 1 𝜕𝜕𝜕𝜕 𝑧𝑧 2 are usually usedof a semicond in theuctor final surface inspection using a common 𝐹𝐹𝑒𝑒𝑒𝑒 𝑧𝑧 − 2 𝑉𝑉𝑐𝑐𝑐𝑐 𝑉𝑉𝑑𝑑𝑑𝑑 𝑉𝑉𝑎𝑎𝑎𝑎𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 𝜕𝜕𝜕𝜕 surfaces after contact electrification, using (4.0) 3 touch-up tool and propose a model for the Thus, potential map images have two process of manufacturing, prior to shipment. charge decay at the surface of Si sample Therefore, it is in the interest of this work to components: tip bias ( and ) and capacitance after CE. 3 Thus, potential map images have two study the effects of contact electrification of a gradient . Thus, capacitance gradient is components: tip bias ( and ) and semiconductorK surfaceFM using a common touch- influenced by the surface charge density on the capacitance gradient ( ( )) . Thus, 𝑎𝑎𝑎𝑎 𝑑𝑑𝑑𝑑 up tool and propose a model for the charge surfacecapacitance of the gradient sample. is Therefore,𝑉𝑉influenced𝑉𝑉 by the the change Surface potential microscopy or SPM surface charge density on𝜕𝜕𝜕𝜕 the𝑧𝑧 surface⁄ 𝜕𝜕𝜕𝜕 of the decay at the surface(Melitz, ofShen, Si sample Kummel, after & Lee CE., 2011) is a sample. Therefore, the change in surface special function of an AFM, and it is also charge density changes the force on the tip known commonly as KFM, which produces ( ) and therefore changes the potential images of potential maps from the scanned image of the surface. surface. The cantilever tip function is 𝐹𝐹𝑒𝑒𝑒𝑒 𝑧𝑧 Commented [LJA2]: Please check if changes reflect the related to the energy of the capacitance C intended meaning. OK between the tip and the sample. The AFM commonly uses a tapping mode to electrostatic force ( ) between the tip scan the surface (“Basic Theory Atomic Force Microscopy (AFM),” n.d.) to obtain its and sample surface is then related to the 𝑒𝑒𝑒𝑒 physical topography. In this method, the tip rate of changes of energy𝐹𝐹 𝑧𝑧 that lightly touches the surfaces (thus tapping)

( ) at its oscillating frequency (60 to 100 kHz) ( ) = ( ) (1.0) while maintaining constant oscillation 1 𝑑𝑑𝑑𝑑 𝑧𝑧 2 amplitude to generate the image. The 𝐹𝐹𝑒𝑒𝑒𝑒 𝑧𝑧 − 2 𝑑𝑑𝑑𝑑 ∆𝑉𝑉 where is the voltage difference between tapping mode produces high-resolution tip and sample. Where ( ( )) is the imaging without leaving artefact or damage capacitance∆𝑉𝑉 gradient with respect to z, on the surface of the samples. The which is the vertical 𝜕𝜕𝜕𝜕distance𝑧𝑧 ⁄ 𝜕𝜕𝜕𝜕between topographic image generated by tapping sample and tip. Whereas is defined as mode AFM is often called height image.

= ( ) + + ∆𝑉𝑉 (2.0) SPM on the other hand utilizes the same tapping initially to generate height images. 𝑎𝑎𝑎𝑎 𝑑𝑑𝑑𝑑 𝑐𝑐𝑐𝑐 where∆𝑉𝑉 𝑉𝑉 and𝑡𝑡 𝑉𝑉 ( ) 𝑉𝑉are the DC bias Height images are stored, and the tip then voltage on the tip, ( ) = is the AFM lifts up to a certain height point called z 𝑑𝑑𝑑𝑑 𝑎𝑎𝑎𝑎 oscillating𝑉𝑉 drive voltage𝑉𝑉 𝑡𝑡 on tip and is the height. Utilizing the stored image from the 𝑎𝑎𝑎𝑎 resonant frequency𝑉𝑉 of𝑡𝑡 the𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 cantilever𝑠𝑠 , and previous tapping mode scan, the entire is contact potential due to work function𝜔𝜔 surface is scanned again at a fixed difference between tip ( ) and sample amplitude set point AO, which followed the 𝑐𝑐𝑐𝑐 and𝑉𝑉 is expressed as potential gradient of the surface. This 𝑡𝑡 𝑠𝑠 ∅ ∅ second image is then interleaved with the ( ) = (3.0) first image to generate surface potential ∅𝑠𝑠−∅𝑡𝑡 image as shown in Figure 1. 𝑉𝑉𝑐𝑐𝑐𝑐 𝑞𝑞 q is the electronic charge. Substituting (3.0) to (2.0) and then to (1.0), therefore, the Figure 1. Concept of AFM-KFM. Commented [LJA3]: Insert Figure 1 here.

electrostatic force ( ) exerted between The SPM (KFM) imaging techniques the tip and sample can be represented as 𝑒𝑒𝑒𝑒 used in this experiment will give some (Melitz et.al., 2011) 𝐹𝐹 𝑧𝑧 insights into changes of the material

surfaces after contact electrification, using

SURFACE ANALYSIS OF SILICON (100) WAFER ESMERIA & POBRE 117 in surface charge density changes the force on MATERIALS AND METHODS the tip and therefore changes the potential image of the surface. A 5-mm × 5-mm n-type Si wafer (100) (SUMCO AFM commonly uses a tapping mode Corporation) was used as a sample for this to scan the surface (“Basic Theory Atomic experiment. Using conventional AFM function, Force Microscopy (AFM),” n.d.) to obtain its the silicon wafer flatness was measured to be physical topography. In this method, the tip 0.12 nm ± 0.2 nm for a scan length of 40 µm lightly touches the surfaces (thus tapping) at with an average surface roughness of 0.26 its oscillating frequency (60 to 100 kHz) while nm. The sample was cleaned using a similar maintaining constant oscillation amplitude type of cotton Q-tip soaked with 2-propanol; to generate the image. The tapping mode its surface was wiped in one direction and produces high-resolution imaging without then left to dry at room temperature for 300 leaving artefact or damage on the surface of seconds. This procedure was to mitigate the the samples. The topographic image generated effects of contamination during 2D scanning. by tapping mode AFM is often called height The sample was attached to a conductive image. carbon double-sided tape on a 3-mm × 1-mm SPM on the other hand utilizes the same stainless steel disk. tapping initially to generate height images. A cleanroom-grade industrial cotton Q-tip Height images are stored, and the tip then lifts (HUBY340) («3» Cotton Applicator |Clean up to a certain height point called z height. Cross,» n.d.) was used as surface applicator on Utilizing the stored image from the previous the Si wafer to generate contact electrification. tapping mode scan, the entire surface is Figure 3 shows the type of cotton Q-tip used scanned again at a fixed amplitude set point in these experiments and the SEM image of

AO, which followed the potential gradient the cotton fiber at 10,000×, showing a small of the surface. This second image is then particle encircled in the picture. The swab interleaved with the first image to generate was rubbed more than five times, and each surface potential image as shown in Figure 1. rub has an equivalent down force of 100-g The SPM (KFM) imaging techniques used weight pressed on the surface of the Si (100) in this experiment will give some insights into wafer. The Si is scanned initially to image changes of the material surfaces after contact the uncharged surface, then the sample was electrification, using a practical commonFIGURES touch ANDretracted TABLES from the AFM to perform contact in manufacturing such as the cotton Q tip. electrification. Field of view is set at 1µm × 1

Figure 1. Concept of AFM-KFM. Figure 1. Concept of AFM-KFM, The tapping mode (1) scans the surface of the sample to obtain the topography of the surface (left) and stores it. The tip lifts up (lift mode), in (2) to z height value, and rescans the entire surface. The images are interleaved so that surface potential is produced along with the topographic data (right).

Figure 2. Groove marker on the surface of Si (100) wafer. The surface of the Si (100) wafer the groove marker (white square) as seen through the AFM optical scope. This marker aids in scanning the same location during SPM imaging. Inset picture (500×) shows the AFM tip. Figure 3. Cotton swab fiber magnified 10,000× using FE-SEM. An example of a Q-tip (left) used in rubbing the surface of the Si (100) wafer and its fiber magnified (yellow box) at 10,000× using FE-SEM. Yellow circle indicates foreign material or contaminant.

118 MANILA JOURNAL OF SCIENCE VOLUME 10 (2017)

µm with an image aspect ratio of 1:1. A grooved All experiments were done in a laboratory at marker was utilized to scan the exact location a room temperature of 22±4°C and a relative of the scanned section of the Si wafer after humidity (RH) of 50±5%. contact electrification as shown in Figure 2. A commercially available Bruker Dimension ICON Atomic Force MicroscopeFIGURES AND TABLESRESULTS AND DISCUSSION (AFM) with Surface Potential Microscopy (SPM) modes was utilized for the experiment. AFM Topographic and KFM Surface

The AFM resolution is 0.01 nm, and the tips Potential Images used were commercially available n-type MESP-RC silicon tips with CoCr coating Figures 4A and 4B show the topographic from Bruker (“AFM Probes | AFM Cantilever (height) images in nanometers using AFM with | AFM Tips - Bruker AFM Probes,” n.d.). It a scan size of 1 µm × 1 µm, before and after has a nominal radius of about 20 nm, and a CE, respectively. This height image serves as tip resonant frequency ranging from 70 kHz reference before SPM imaging. It is noted that to 80 kHz was used for both imagings. The the surface of Si after does not significantly tip was biased by about +1 V in order to set change, except for some bright protruding dots. This change in surface texture is a result of the appropriateFigu rdynamice 1. Concept range of AFM on-KFM, the The images. tapping mode (1) scans the surface of the sample to Height imagesobtain were the topography used as of the reference surface (left for) and storeswhen it. The cotton tip lifts and up (lift Si mode),surface in (2) make to contact or both modalitiesz height to value, determine and rescans if the entire changes surface. Theare images rubbed are interleaved against eachso that other. surface The bright dots potential is produced along with the topographic data (right). in the surface structures are created by in Figure 4B are indicative of foreign material surface defects or from airborne materials. left behind during CE.

Figure 2. Groove marker on the surface of Si (100) wafer. The surface of the Si (100) wafer the grooveFigure marker 2. (whiteGroove square) marker as seenon Si through (100) waferthe AFM surface. optical scope. This marker Figureaids in scanning3. Cotton the swab same fiber location magnified during SPM 10,000 imaging.× using Inset FE-SEM. picture (500×) shows the AFM tip. Figure 3. Cotton swab fiber magnified 10,000× using FE-SEM. An example of a Q-tip (left) used in rubbing the surface of the Si (100) wafer and its fiber magnified (yellow box) at 10,000× using FE-SEM. Yellow circle indicates foreign material or contaminant. a) BEFORE CE Topographic image b) AFTER CE Topographic Image

A B

Figure 4. AFM images of n-type Si. Figure 4. AFM Images of n-type Si. (A) AFM HEIGHT (nm) images of n-type Si before CE. (B) The same spot on the surface after application of CE. There is no significant change on the surface, except for bright spots, which are foreign materials left behind during CE with the cotton Q-tip.

c) BEFORE CE SPM Image d) AFTER CE (CHARGE) SPM Image

C D

e) Zoomed in SIDE view of image d) as shown by the arrow

Figure 5. Si surface potential images of n- type. (C) 3D SPM images of n-type Si before contact electrification. (D) The same sample surface after being rubbed with a cotton swab. The “carpet-like” appearances of the surface after contact electrification is observed. (E) Zoomed in side view of the spikes indicated by the arrow in D.

SURFACE ANALYSIS OF SILICON (100) WAFER ESMERIA & POBRE 119

Using SPM imaging the surface before in the electronic industries as a touch-up tool CE is shown in Figure 3C together with to minimize damage on electronic devices. To the charged surface in Figure 4D after CE. compare the surface potential image of the The surface before CE shows a mosaic but Si (100) wafer after CE, two insulator-type smooth pattern, whereas the surface after CE materials, namely, muscovite mica (KAl2 image shows a carpet-like surface, wherein (SI3Al)O10(OH)2) and silicone rubber were individual spikes distance between each other used to compare with cotton. Silicone rubber, is approximately 17 nm ± 10 nm, based from which is based from polydimethylsiloxane the image sectional analysis as shown in (PDMS), has a negative (−) charge affinity, Figure 6. The mosaic-like pattern is images of while muscovite mica has a positive (+) charge charge distribution on the surface before CE; affinity in the triboelectric series (Lacks & after CE, the carpet-like image in the surface Mohan Sankaran, 2011). appeared, which is composed of positive (light Using a silicone (Fuji Silicone) rubber, color) and negative (dark color), as shown in CE was performed by rubbing the Si surface the inset image. five times in one direction, with a downward KFM images were also taken using force approximately at 300 g. Figures 7C and 5-µm × 5-µm scan size using the same type 7D show the similar carpet-like images in the of cotton Q-tip and the same methodology figure 5d Si (100) wafer. Since silicone negative for generating CE. Figures 7A and 7B show charge affinity is lower than cotton in the the Si (100) wafer that manifested the same triboelectric series, the surface potential image carpet-like image similara) BEFORE to Figure CE Topographic 5 with image peak b)would AFTER CE have Topographic a higher Image positive spike intensity distances of approximatelyA 13 nm ± 5 nm. comparedB to cotton in Figure 7b. Each peak distance was approximately 12 ± 5 nm. The KFM Image Comparison of Cotton, muscovite mica used for applying CE was a Muscovite Mica, and Silicone Rubber (Muscovite Mica Substrates: SPI Supplies, n.d.) disk with a 12-mm diameter and 0.15-mm

Cotton is reported to be neutral, or its charge thickness. This disk was placed on top of the affinity is very small inFig uthere 4. AFMtriboelectric Images of n-type Siseries. (A) AFM HEIGHTSi samples (nm) images andof n-type pressed Si before CE using. a 200-g stainless (B) The same spot on the surface after application of CE. There is no significant change on (Diaz & Felix-Navarro,the surface, 2004). except for This bright spots, is thewhich are foreignsteel materials weight left behind for duringabout CE with600 seconds. the cotton Q-tip. reason why cotton Q-tips are commonly used

c) BEFORE CE SPM Image d) AFTER CE (CHARGE) SPM Image

C D

e) Zoomed in SIDE view of image d) as shown by the arrow

Figure 5. Si surface potential images of n- type. (C) 3D SPM images of n-type Si before contact electrification. (D) The same sample surface after being rubbed with a cotton swab. The “carpet-like” appearances of the surface after contact electrification is observed. (E) Figure 5. Si KFMZoomed images in side view of ofthe the spSiikes (100) indicated wafer by the arrowand inside D. view of the carpet-like images showing positive and negative spikes. 12

120 MANILA JOURNAL OF SCIENCE VOLUME 10 (2017)

a) TOP VIEW 2D image of the surface b) Plot of the line section (yellow line). potential map

a) Cotton b)

Cotton a) b)

Figure 6. Line sectionFigure of 6.the Line SPM section image (2D) of .the Left SPM figure image shows the(2D). plot of the line section (yellow arrow) from the SPM 2D image. The line section was plotted on the left (b). a)Average Cotton distance between peaks was 17 nm ± 10 nm. b)

Figure 7. Si surface after CE using cotton q tip, Silicone rubber and muscovite mica sheet. (A) Si(100) wafer rubbed with cotton but with a larger 5-µm × 5-µm scan size. (B) Side view of A

Fshowiguringe 7 .the Si surfacepotential after spikes CE. using cotton q tip, Silicone rubber and muscovite mica sheet. (A)

Si(100) wafer rubbed with cotton but with a larger 5-µm × 5-µm scan size. (B) Side view of A showc) ingSilicone the potential Rubber spikes . d)

Silicone Rubber Fc)igu re 7. Si surface after CE using cotton q tip, Siliconed) rubber and muscovite mica sheet. (A)

Si(100) wafer rubbed with cotton but with a larger 5-µm × 5-µm scan size. (B) Side view of A

showing the potential spikes.

Silicone Rubber c) d)

(C) SPM image of Si after CE using silicone rubber as applicator. (D) Side view of the potential spikes; note the relative higher intensity due to larger frictional contact area.

potentiale) spikes; note the relative higher intensity duef) to larger frictional contact area.

Muscovite Mica e) f)

(C) SPM image of Si after CE using silicone rubber as applicator. (D) Side view of the potential spikes; note the relative higher intensity due to larger frictional contact area.

Muscovite Mica e) 13 f)

Figure 7. (E)KFM Potential image map of imageSi (100) of Si after after CEapplication using ofcotton CE using (5 µm muscovite × 5 µm mica), mica,. (F) Side and view silicone of rubber. the potential spikes, which showed tendency to the negative side, which is due to the positive affinity of mica at the triboelectric series. Figures 7E(E) Potentialand 7F map show image the of KFMSi after imageapplication ofDecay CE using Time muscovite mica. (F) Side view of the potential spikes, which showed tendency to the negative side, which is due to the positive of the Si sampleaffinity after of m icabeing at the applied triboelectric CE series. using mica sheet. Mica has higher positive (+) charge It was reported for polymers that there is 14 affinity compared (E) Potential to map cotton; image of SiSi after on application the other ofa CE minimal using muscovite average mica reduction. (F) Side view after of 2.2 hours hand is negativethe potential (−), sospikes, the which resulting showed tendency KFM to thefor negative insulator-like side, which polymersis due to the (Baytekinpositive et. al., affinity of mica at the triboelectric series. 14 image has shown negative potential peaks as 2011). In order to understand the behavior shown in Figures 7E and 7F. of the surface charge in Si after extended

SURFACE ANALYSIS OF SILICON (100) WAFER ESMERIA & POBRE 121 periods of time, surface charge retention was another ~980 seconds. Figure 8 (A to E) shows taken after grounding the sample. The sample the sequence of surface charge images by SPM. was placed on a stainless steel disk fixture 30 In Figure 8A, the initial image is recorded mm in diameter and 3 mm in thickness and after 50 seconds. Minimal change had been was attached to conductive carbon tape with observed on the average surface potential from surface resistance of <105 Ω. The setup allows A to B; in C, the image shows some spikes the sample to be electrically grounded from the appearing along the edges of the surfaces, fixture to AFM base plate. This setup allows but subsequently, these spikes are reduced as the sample to be continuously discharged shown in D (after 550 seconds). Both positive during the acquisition of the image in the and negative surface charge distributions on SPM tapping mode. Contact electrification was Si (100) wafer have dissipated in the order performed again using the same method as of ±2.5 mV after 4800 seconds of noncontact previously described. SPM scanning duration electrification. took about 1,080 seconds (~0.3 hour) from Figure 8E shows the potential profile contact electrification, which includes tip with very small texture, suggesting the engagement; loading on the sample ~100 surface potential has decayed to 0. These seconds; and actual scanning, which takes observations are similar to the previous work

a) b)

d c) )

Figure 8. Surface potential profile and decay time. SPM surface images of the Si (100) wafer after the electrostatic charge was allowed to dissipate through grounded fixture. Figure 8. Surface potential profile and decay time.

122 MANILA JOURNAL OF SCIENCE VOLUME 10 (2017) on CE showing that Si (111) is stable after 800 capacitor are connected in series as shown seconds (Shiota, n.d.), which indicated a long in Figure 10. It can deduce that the surface relaxation time. charge generated from the triboelectrification The average surface height deviation was behaves like a capacitor-like element while the used to measure changes in surface potential Si wafer–carbon conductive junction behaves value of Figures 8A to 8E. These values are like a diode. referred to as average roughness (Ra) using Thus, the equivalent D.C. electric circuit NanoScope software ver. 11 of the Bruker can be treated like a clamped capacitor circuit ICON AFM wherein the image selected where the surface charge is being discharged through a cursor box calculates this value in by the n-type Si/carbon conductive tape millivolts. Schottky diode-like junction with a constant Figure 9 (A to E) shows the temporal positive voltage in equation (2.0), which is surface potential (in millivolts) of the Si the breakdown voltage of the Si wafer and wafer even after a period of 4800 seconds. carbon conductive tape junction. Thus, the The surface potentials are derived from the decay time constant of the uniformly surface inset SPM images that were averaged from charge distribution in the capacitor like on the entire surface area with height deviations. the Si wafer can be described in terms of the The dynamics of the surface charge on potential difference between the cantilever the Si (100) wafer brought about by the and the grounded steel disk as Panofsky and triboelectrification can be modeled like a Phillips (1978, p .123) show in Figure 10. D.C. electric circuit where the diode and a / . 60 ( ) = 14.28 + 48.50 𝑡𝑡 4334 79 b 𝑉𝑉 𝑡𝑡 𝑒𝑒 45 a / . 60 ( ) = 14.28 + 48.50 𝑡𝑡 4334 79 b 𝑉𝑉 𝑡𝑡 𝑒𝑒 c Predicted 30 d V(t) mV 45 a Data e 15 c Predicted 30 d V(t) mV Data 0 e 15 10 100 1000 10000 Log t (Seconds)

0 Figure 9:. Plot of Si decay time. Plot of V(t) against Log of relaxation time t (second). Letters A to E corresponds10 to the images 100in Figures 10A1000 to 10E. The g10000reen line is the predicted Log t (Seconds) , model derived from the raw data points with extracted and t values.

𝑂𝑂 𝑆𝑆 Figure 9:. Plot of Si decay timeFigure. Plot of9. V Plot(t) against of Si Logdecay of𝑉𝑉 relaxation time.𝑉𝑉 time t (second). Letters A to E corresponds to the images in Figures 10A to 10E. The green line is the predicted model derived from the raw data points with extracted , and t values.

𝑂𝑂 𝑆𝑆 cantilever 𝑉𝑉 𝑉𝑉 +

surface charge 𝑡𝑡 − ( ) 0 𝜏𝜏 n-Si 𝑉𝑉 𝑒𝑒 cantilever − + 𝑉𝑉 𝑡𝑡 𝑉𝑉𝑠𝑠 Carbonsurface tape charge 𝑡𝑡 − 𝜏𝜏 ( ) Steel disk 0 n-Si 𝑉𝑉 𝑒𝑒 −

𝑠𝑠 𝑉𝑉 𝑡𝑡 Figure 10Carbon. Equivalent tape D.C. circuit model. Equivalent D.C. electric𝑉𝑉 circuit model of the cantilever-surface charge.Figure An 10. air Equivalentgap capacitor -D.C.like element circuit connected model. to an n-type Si/carbonSteel conductive disk tape Schottky diode-like element connected to the grounded steel disk.

F igure 10. Equivalent D.C. circuit model. Equivalent D.C. electric circuit model of the cantileverTable 1. List-surface of P charge.ossible A Dn epletionair gap capacitor Thickness-like delementSi (nm) ofconnected Si Involved to an inn-type Discharging Si/carbonthe conductive Surface C tapeharge Schottky Given diode Two- likePossible element Input connected Parameters to the grounded of ρair and steel dai disk.r

Air Resistivity, ρ air (Ω -m) Air Thickness, dair (nm) Depletion Thickness, dsi (nm)

Table 1. List of Possible Depletion Thickness8 dSi (nm) of Si Involved27 in Discharging 16 the Surface1.3 × 10 Charge Given Two Possible Input Parameters of ρair and dair 16 5 Air Resistivity, ρ air (Ω -m) Air Thickness, dair (nm) Depletion Thickness, dsi (nm) 3.3 × 1017 8 70 8 27 1.3 × 1016 16 16 5 3.3 × 1017 8 70

16 retention was taken after grounding the (Ra) using NanoScope software ver. 11 of sample. The sample was placed on a the Bruker ICON AFM wherein the image stainless steel disk fixture 30 mm in selected through a cursor box calculates diameter and 3 mm in thickness and was this value in millivolts. attached to conductive carbon tape with surface resistance of <105 Ω. The setup Figure 9 (A to E) shows the temporal Commented [LJA14]: Please check if changes reflect the intended meaning. OK allows the sample to be electrically surface potential (in millivolts) of the Si grounded from the fixture to AFM base wafer even after a period of 4800 seconds. plate. This setup allows the sample to be The surface potentials are derived from the continuously discharged during the inset SPM images that were averaged from acquisition of the image in the SPM tapping the entire surface area with height mode. Contact electrification was performed deviations. again using the same method as previously described. SPM scanning duration took Figure 9. Plot of Si decay time. Commented [LJA18]: Please insert Figure 9 here. about 1,080 seconds (~0.3 hour) from contact electrification, which includes tip The dynamics of the surface charge on engagement; loading on the sample ~100 the Si (100) wafer brought about by the seconds; and actual scanning, which takes triboelectrification can be modeled like a another ~980 seconds. Figure 8 (A to E) D.C. electric circuit where the diode and a shows the sequence of surface charge capacitor are connected in series as shown images by SPM. In Figure 8A, the initial in Figure 10. It can deduce that the surface image is recorded after 50 seconds. Minimal charge generated from the Commented [LJA15]: Please check if changes reflect the change had been observed on the average triboelectrification behaves like a capacitor- intended meaning. OK surface potential from A to B; in C, the like element while the Si wafer–carbon image shows some spikes appearing along conductive junction behaves like a diode. the edges of the surfaces, but subsequently, these spikes are reduced as shown in D Figure 10. Equivalent D.C. circuit model. Commented [LJA19]: Insert Figure 10 here. (after 550 seconds). Both positive and negative surface charge distributions on Si Thus, the equivalent D.C. electric circuit (100) wafer have dissipated in the order of can be treated like a clamped capacitor ±2.5 mV after 4800 seconds of noncontact circuit where the surface charge is being electrification. discharged by the n-type Si/carbon conductive tape Schottky diode-like junction Figure 8. Surface potential profile and with a constant positive voltage in equation (2.0), which is the breakdown Commented [LJA16]: Insert Figure 8 here. decay time. 𝑠𝑠 voltage of the Si wafer and carbon𝑉𝑉 Figure 8E shows the potential profile conductive tape junction. Thus, the decay with very small texture, suggesting the time constant of the uniformly surface surface potential has decayed to 0. These charge distribution in the capacitor like on observations are similar to the previous the Si wafer can be described in terms of work on CE showing that Si (111) is stable the potential difference between the after 800 seconds (Shiota, n.d.), which cantilever and the grounded steel disk as SURFACE ANALYSIS OF SILICON (100) WAFER ESMERIA & POBRE 123 indicated a long relaxation time. Panofsky and Phillips (1978, p .123) show Commented [LJA17]: Please check if changes reflect the in Figure 10. intended meaning. OK 16 14 The average surface height deviation was air-Si interface layer for the capacitor model The electric potential (() )measured by used to measure changes in surface The electric( ) = potential+ measured (5.0) by is shown in Figure 11. 𝑡𝑡 (5.0) Insert Figure 11 here. the surface potential mode of AFM is the Figure 11. Depletion region. CommentedCommented [LJA20]: [LJA20]: Insert Figure 11 here. potential value of Figures 8A to 8E. These the surface potential−𝜏𝜏 mode of AFM is the Figure 11. Depletion region. total surface 0potential𝑠𝑠 from𝑉𝑉 𝑡𝑡 the charging values are referred to as average roughness total surface𝑉𝑉 𝑡𝑡 𝑉𝑉potential𝑒𝑒 𝑉𝑉 from𝑉𝑉 𝑡𝑡 the charging capacitorThe electric-like potentialelement withmeasured an byinitial the Substituting the electric permittivities dair air capacitorsurface potential-like element mode of AFMwith is an the totalinitial Substituting theair electric permittivities surface potential of from anddSi resisitivitiesSi of bothSi air and silicon surface potential of from and Si resisitivities of both air and silicon 6 surfacetriboelectrification potential from and the acharging constant capacitor- positive materials, the depletioncharge region thickness on air taken tribolikeelectrification element with an and initial a surfaceconstant𝑉𝑉0 potential positive materials,C - tape the charge thickness on air taken potential due to the breakdown voltage0 of from SPM image ( = 12 ± 4 ), the potentialofth e from n- type triboelectrificationdue toSi /thecarbon breakdown conductive and 𝑉𝑉 avoltage constant tape of decayfrom time SPM constantimage ( from= 12the fitted± 4 ), the 𝑎𝑎𝑎𝑎𝑎𝑎 thpositiveeSchottky n-type potentialjunction Si/carbon. dueIn equation to conductive the breakdown (2.0) , tapeis exponentialdecay timegreen line𝑑𝑑constant from F𝑛𝑛𝑛𝑛 igurefrom 𝑛𝑛𝑛𝑛6, whichthe fitted Figure 11. Depletion𝑎𝑎𝑎𝑎𝑎𝑎 region. Schottkyvoltageconsidered ofjunction thethe n-typedecay. In Si/carbontime equation constant conductive (2.0) ofFig u,ther e 1 1 .is Illustration is exponential= showing 4334.79 the depletion secondsgreen region ,line. 𝑑𝑑Theinto d epletionfrom equation regionF𝑛𝑛𝑛𝑛igure between ( 4𝑛𝑛𝑛𝑛6) ,the which n- consideredtapeair between Schottky the the junction. decay Si wafer timeIn andequation constant the cantilever (2.0),type of𝜏𝜏 Siis thewafer andwith theis acarbon potential= conductive4334.79 barriertape givesseconds riseof to15.6 the, 15.into mV6-mV forpotentialequation the barrier (4) between the n-type 𝜏𝜏Si and carbon conductive tape given the 48.5-mV surface potential from airconsideredof between the AFM the thesystem. decaySi wafer For time the and constant condition the cantilever of contactset the in𝜏𝜏 electrification n-typewith with Si/a cottonacarbon potential Q-tip. conductive barrier tape of 15.6Schottky mV for the Substituting the electric permittivities and ofair thethe between experiment,AFM system. the Si decay wafer For time theand conditionconstantthe cantilever setcan in dioden--typelike𝜏𝜏 junctionSi/carbon and conductive a surface potentialtape Schottky ofbe the estimated AFM system. as For the condition set in ofresisitivities 48.5 mV, ofwe both can air predict and silicon the materials,depletion the experiment, decay time constant can diode-like junction and a surface potential the experiment, decay time constant can𝜏𝜏 be thicknessthe charge of thickness Si involved on air in takendischarging from SPM the be estimated as= (6.0) surfaceimageof 48.5(), charge the mV decay as, shownwe time can in constant Tablepredict 1 from the thedepletion estimated as 𝜏𝜏 fittedthickness exponential of Si involvedgreen line in dischargingfrom Figure 6, the 𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒 where 𝜏𝜏= is(𝜀𝜀 the𝜌𝜌 ) effective (6.0)(6.0) relative whichsurface is 4334.79 charge T seconds,able as shown 1. into in equation Table 1 (4) Commented [LJA21]: Please insert Table 1 here. permittivity of both air and silicon in Farad with a potential barrier of 15.6 mV for the 𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒 wherewhereper meter is𝜀𝜀 𝜏𝜏 the effectiveis(and𝜀𝜀 the𝜌𝜌 relative) effectiveis the permittivity effectiverelative withn-type two Si/carbon different conductive inputTable parameters tape1. Schottky: the Commented [LJA21]: Please insert Table 1 here. of both air and silicon in Farad per meter resistivitydiode-like junctionof air and and the a air surface gap thickness potential permittivityresistivity ofof bothboth air air𝑒𝑒𝑒𝑒𝑒𝑒 and and silicon silicon in in Ohm Farad- andmeter is .the 𝜀𝜀The𝑒𝑒𝑒𝑒𝑒𝑒 effective relative resistivity𝜌𝜌 electric ofpermittivity both air and of dofair 48.5with taken mVtwo from, we different SPM can predictmeasurements. input the parameters depletion It can : the per meter and is the effective be shown in Table 1 that the thickness of siliconair at STP in Ohm-meter. is = 8.9 Thex 10 relative-12 F/m, electricand the thicknessresistivity of Si of involved air and in the discharging air gap thethickness resistivity of both air and silicon in Ohm- the Si wafer should be more than 70 nm to permittivityrelative permittivity of air at STP𝑒𝑒𝑒𝑒𝑒𝑒 of is =silicon 8.9 x 10is-12 F/m,= surface charge as shown in Table 1 meter. The relative𝑎𝑎𝑎𝑎𝑎𝑎 𝜌𝜌 electric permittivity of providedair takenenough from space SPM for measurements.discharging the It can and11.7 the 10 relative F/m𝜀𝜀 , whilepermittivity the resistivity of silicon of airis -12 𝑆𝑆𝑆𝑆 be shown in Table 1 that the thickness of air at STP is 16= 8.9 x 10 17 F/m, and𝜀𝜀 16 the capacitor in the proposed model. F/m,is while = −1.314 the x resistivity10 ~3.3 ×of 10air isΩ =-m 1.3 and x 10 the Tablethe 1. Si List wafer of Possible should Depletion be more Thickness than 70 nm to 𝑥𝑥 17 = relative~resistivity3.3 × 10permittivity Ωof-m silicon and theis 0.2resistivityof Ωsilicon-m ("Properti of siliconis es dSi (nm) of Si Involved in Discharging the 𝑎𝑎𝑎𝑎𝑎𝑎 𝜀𝜀𝑎𝑎𝑎𝑎𝑎𝑎 provide enough space for discharging the 11isof.7 0.2𝜌𝜌 10Silicon Ω-m F/m («Properties—El,- Cat.com,"while the of Silicon—El-Cat. resistivityand Virginia of air Surface Charge Given Two Possible Input 𝑆𝑆𝑆𝑆 capacitorCONCLUSION in the proposed model. is com,»Semiconductor = and −1.314 Virginia x 10 Inc.,16 Semiconductor~ 3.3n.d) .× For 10 17the Ω -Inc.,compositem and n.d).𝜀𝜀 the Parameters of ρair and dair

Formaterial𝑥𝑥 the composite of air and material silicon ofin air the and capacitor silicon- 17 resistivity of silicon is 0.2 Ω-m ("Properti es The surface charge on the Si (100) wafer like𝜌𝜌𝑎𝑎𝑎𝑎𝑎𝑎 element, with common surface area A, Air Air Depletion ofin theSilicon capacitor-like—El-Cat.com," element, withand commonVirginia was demonstrated using the tapping mode the decay time constant can be calculated Resistivity, Thickness,CONCLUSION Thickness, Semiconductorsurface area A, Inc.,the decay n.d) .time For constant the composite can of KFM. KFM surface images of the Si (100) in terms of the permittivities and ρair (Ω-m) dair (nm) dsi (nm) be calculated in terms of the permittivities and wafer after CE can have either more materialresistivities of air ofand both silicon 𝜏𝜏air in theand capacitorsilicon - 8 27 resistivities of both air and silicon respectively, positive1.3 ×T 1016 orhe negative surface peakscharge depending on the Si on ( 100the ) wafer likerespectively, element, thatwith can common be derived surface as area A, 16 5 that can be derived as chargewas affinitydemonstrated of the material using thein contact.tapping mode the decay time constant can be calculated Also,of itKFM. was determined KFM surface8 that images surface70 of charge the Si (100) in terms of the( permittivities) and 3.3 × 1017 = = (7.0) (7.0) didwafer not diminish after upCE 16to can50% ofhave its 14 originaleither more resistivities of 𝜀𝜀𝑎𝑎𝑎𝑎𝑎𝑎both𝜀𝜀𝑆𝑆𝑆𝑆 𝜌𝜌𝑎𝑎𝑎𝑎𝑎𝑎 𝑑𝑑𝜏𝜏𝑎𝑎𝑎𝑎𝑎𝑎air+𝜌𝜌 𝑆𝑆𝑆𝑆𝑑𝑑𝑆𝑆and𝑆𝑆 silicon surfacepositive charge or negative even after peaks 550 depending seconds. on the respectively,𝜏𝜏 𝜀𝜀𝑒𝑒𝑒𝑒𝑒𝑒𝜌𝜌𝑒𝑒𝑒𝑒𝑒𝑒 that can𝜀𝜀𝑆𝑆𝑆𝑆𝑑𝑑 be𝑎𝑎𝑎𝑎𝑎𝑎+ derived𝜀𝜀𝑎𝑎𝑎𝑎𝑎𝑎𝑑𝑑𝑆𝑆𝑆𝑆 as InIn understandingunderstanding howhow thethe capacitorcapacitor andand Surfacewithcharge two potential differentaffinity distribution inputof the parameters: material on the in theSi contact.

diodediode modelsmodels areare separated,separated, thethe depthdepth ofof (100)resistivityAlso, wafer it ofwas has air determined andrelatively the air small gapthat thicknesspotential surface charge the depletion region between( n-type) Si and changed taken in from the SPMorder measurements. of ±2.5 mV even It can after be the= depletion= region between n-type Si and(7.0) air did not diminish up to 50% of its original carboncarbon conductiveconductive𝜀𝜀𝑎𝑎𝑎𝑎𝑎𝑎 𝜀𝜀 tape𝑆𝑆tape𝑆𝑆 𝜌𝜌𝑎𝑎𝑎𝑎𝑎𝑎 (a𝑑𝑑(a𝑎𝑎𝑎𝑎𝑎𝑎 SchottkySchottky+𝜌𝜌𝑆𝑆𝑆𝑆𝑑𝑑𝑆𝑆𝑆𝑆 diode-diode- 4800shown seconds in Table of 1noncontact that the thickness electrification. of the SiA surface charge even after 550 seconds. likelike element)𝑒𝑒𝑒𝑒𝑒𝑒element)𝑒𝑒𝑒𝑒𝑒𝑒 providesprovides𝜀𝜀𝑆𝑆𝑆𝑆𝑑𝑑 𝑎𝑎𝑎𝑎𝑎𝑎 thethe+ 𝜀𝜀 demarcation𝑎𝑎𝑎𝑎𝑎𝑎demarcation𝑑𝑑𝑆𝑆𝑆𝑆 lineline directwafer shouldcurrent be (D.C.) more thanelectric 70 circuitnm to providemodel, 𝜏𝜏 𝜀𝜀 𝜌𝜌 Surface potential distribution on the Si betweenbetweenIn understanding thethe capacitorcapacitor how andand thethe diode diodecapacitor modelsmodels and whichenough consistspace fors ofdischarging a clamped the capacitor capacitor in (100) wafer has relatively small potential diode models are separated,Si the depth of series with a Schottky diode, explains how wherewhere thethe chargecharge thicknessthickness ddSi ofof SiSi onon thethe the proposed model. theair depletion-Si interface region layer between for the n- typecapacitor Si and thechange surface inpotential the order of ofthe ± 2.5Si wafermV even is after carbonmodel conductiveis shown in Ftapeigure (a11 . Schottky diode- slowly4800 diminished seconds of with noncontact a time constant electrification. of A like element) provides the demarcation line direct current (D.C.) electric circuit model, between the capacitor and the diode models7 which consists of a clamped capacitor in where the charge thickness dSi of Si on the series with a Schottky diode, explains how air-Si interface layer for the capacitor the surface potential of the Si wafer is model is shown in Figure 11. slowly diminished with a time constant of

124 MANILA JOURNAL OF SCIENCE VOLUME 10 (2017)

CONCLUSION Degradation of GMR and TMR recording heads using very short duration ESD transients. IEEE The surface charge on the Si (100) wafer was International Digest of Technical Papers on demonstrated using the tapping mode of KFM. Magnetics Conference 38(5),2283-2285. KFM surface images of the Si (100) wafer after Baytekin, B., Baytekin, H. T., & Grzybowski, B. A. (2012). What really drives chemical reactions CE can have either more positive or negative on contact charged surfaces? Journal of the peaks depending on the charge affinity of the American Chemical Society, 134(17), 7223–7226. material in contact. Also, it was determined Baytekin, H. T., Patashinski, A. Z., Branicki, M., that surface charge did not diminish up to 50% Baytekin, B., Soh, S., & Grzybowski, B. A. of its original surface charge even after 550 (2011). The mosaic of surface charge in contact seconds. Surface potential distribution on the electrification.Science , 333(6040), 308–312. Si (100) wafer has relatively small potential Basic theory atomic force microscopy (AFM) change in the order of ±2.5 mV even after 4800 [pdf]. (n.d.). Retrieved from http://asdlib. seconds of noncontact electrification. A direct org/onlineArticles/ecourseware/Bullen/ current (D.C.) electric circuit model, which SPMModule_BasicTheoryAFM.pdf. consists of a clamped capacitor in series with Brunkov, P. N., Goncharov, V. V., Rudinsky, M. E., a Schottky diode, explains how the surface Gutkin, A. A., Gordeev, N. Y., Lantratov, V. M.,… potential of the Si wafer is slowly diminished Konnikov, S. G. (2013). Local triboelectrification of an n-GaAs surface using the tip of an atomic- with a time constant of 4334.79 seconds and a force microscope. Semiconductors, 47(9), breakdown voltage of 15.6 mV at the depletion 1170–1173. region of Si/carbon tape interface. Cai, W., & Yao, N. (2016). Dynamic nano- triboelectrification using torsional resonance mode atomic force microscopy. Scientific ACKNOWLEDGEMENTS Reports, 6(1). Diaz, A., & Felix-Navarro, R. (2004). A semi- The authors would like to give their sincerest quantitative tribo-electric series for polymeric appreciation to Misses Shirley Galicia and Allana materials: The influence of chemical structure Lopez for their assistance during AFM/ SPM and properties. Journal of Electrostatics, 62(4), setup. Also, the corresponding author would like 277–290. to thank Messrs. Hidehiro Sakurai and Arun El-Mouloud Zelmat, M., Rizouga, M., Tilmatine, A., Khanna, for allowing the corresponding author Medles, K., Miloudi, M., & Dascalescu, L. some AFM time at TPC Reliability Engineering (2013). Experimental comparative study of Laboratory. different tribocharging devices for triboelectric separation of insulating particles. IEEE Transactions on Industry Applications, 49(3), REFERENCES 1113–1118. Gady, B., Reifenberger, R., & Rimai, D. S. (1998). 3” Cotton Applicator|Clean Cross. (n.d.). Retrieved Contact electrification studies using atomic from http://www.cleancross.net/english/ force microscope techniques. Journal of Applied products/threeinch_standard.html Physics, 84(1), 319–322. AFM Probes | AFM Cantilever | AFM Tips— KAB Electro Acoustics. (n.d.). Retrieved from http:// Bruker AFM Probes. (n.d.). Retrieved from www.kabusa.com http://www.brukerafmprobes.com/ Hogue, M. (2004). Insulator/insulator contact Bailey, A. G. (2001). The charging of insulator charging and its relationship to atmospheric surfaces. Journal of Electrostatics, 51–52, pressure . Journal of Electrostatics, 61(3–4), 82–90. 259–268. Baril, L., Nichols, M., & Wallash, A. (2002). Lacks, D. J., & Mohan Sankaran, R. (2011). Contact SURFACE ANALYSIS OF SILICON (100) WAFER ESMERIA & POBRE 125

electrification of insulating materials. Journal Sakaguchi, M., Makino, M., Ohura, T., & Iwata, T. of Physics D: Applied Physics, 44(45), 453001. (2014). Contact electrification of polymers due Mazur, T., & Grzybowski, B. A. (2017). Theoretical to electron transfer among mechano anions, basis for the stabilization of charges by radicals mechano cations and mechano radicals. Journal on electrified polymers.Chemical Science, 8(3), of Electrostatics, 72(5), 412–416. 2025–2032. Shiota. Retrieved from www.oetg.at/fileadmin/ Melitz, W., Shen, J., Kummel, A. C., & Lee, S. Dokumente/oetg/Proceedings/.../M-11-10-658- (2011). Kelvin probe force microscopy and its SHIOTA.pdf application. Surface Science Reports, 66(1), Tipikin. ( 2002) Article retrieved from www. 1–27. tipikindmitriy.hostoi.com/web_documents/ Muscovite Mica Substrates: SPI Supplies. (n.d.). part1.pdf Retrieved from http://www.2spi.com/category/ Virginia Semiconductor, Inc. (n.d.). Retrieved from mica-substrates/substrates/ http://www.virginiasemi.com Németh, E., Albrecht, V., Schubert, G., & Simon, F. Williams, M. W. (2012). Triboelectric charging of (2003). Polymer tribo-electric charging: insulating polymers–some new perspectives. dependence on thermodynamic surface AIP Advances, 2(1), 010701. properties and relative humidity. Journal of Zhou, Y. S., Li, S., Niu, S., & Wang, Z. L. Electrostatics, 58(1–2), 3–16. (2016). Effect of contact- and sliding-mode Panofsky, W. K., & Phillips, M. (1978). Classical electrification on nanoscale charge transfer electricity and magnetism. Mineola, NY: Dover for energy harvesting. Nano Research, 9(12), Publications. 3705–3713. Part 5: Device Sensitivity and Testing » EOS/ESD Zhou, Y. S., Liu, Y., Zhu, G., Lin, Z., Pan, C., Jing, Q., Association, Inc. (n.d.). Retrieved from https:// & Wang, Z. L. (2013). In situ quantitative study www.esda.org/about-esd/esd-fundamentals/ of nanoscale triboelectrification and patterning. part-5-device-sensitivity-and-testing/ Nano Letters, 13(6), 2771–2776. Properties of Silicon—El-Cat.com. (n.d.). Retrieved from http://www.el-cat.com/silicon-properties. htm