molecules
Article Frequency-Dependent Sonochemical Processing of Silicon Surfaces in Tetrahydrofuran Studied by Surface Photovoltage Transients
Artem Podolian 1, Andriy Nadtochiy 1, Oleg Korotchenkov 1 and Viktor Schlosser 2,*
1 Faculty of Physics, Taras Shevchenko National University of Kyiv, 01601 Kyiv, Ukraine; [email protected] (A.P.); [email protected] (A.N.); [email protected] (O.K.) 2 Department of Electronic Properties of Materials, Faculty of Physics, University of Vienna, A-1090 Wien, Austria * Correspondence: [email protected]; Tel.: +43-1-4277-72-611
Abstract: The field of chemical and physical transformations induced by ultrasonic waves has shown steady progress during the past decades. There is a solid core of established results and some topics that are not thoroughly developed. The effect of varying ultrasonic frequency is among the most beneficial issues that require advances. In this work, the effect of sonication of Si wafers in tetrahydrofuran on the photovoltage performance was studied, with the specific goal of studying the influence of the varying frequency. The applied ultrasonic transducer design approach enables the construction of the transducer operating at about 400 kHz with a sufficient sonochemical efficiency. The measurements of the surface photovoltage (SPV) transients were performed on p-type Cz-Si(111)
wafers. Sonication was done in tetrahydrofuran, methanol, and in their 3:1 mixture. When using tetrahydrofuran, the enhanced SPV signal (up to ≈80%) was observed due to increasing sonication ≈ Citation: Podolian, A.; Nadtochiy, A.; frequency to 400 kHz. In turn, the signal was decreased down to 75% of the initial value when the Korotchenkov, O.; Schlosser, V. frequency is lowered to 28 kHz. The addition of methanol suppressed this significant difference. It Frequency-Dependent Sonochemical was implied that different decay processes with hydrogen decomposed from tetrahydrofuran could Processing of Silicon Surfaces in be attempted to explain the mechanism behind the observed frequency-dependent behavior. Tetrahydrofuran Studied by Surface Photovoltage Transients. Molecules Keywords: power ultrasound; sonochemistry; ultrasonic frequency; silicon; tetrahydrofuran; sur- 2021, 26, 3756. https://doi.org/ face photovoltage 10.3390/molecules26123756
Academic Editor: Gregory Chatel 1. Introduction Received: 11 May 2021 It is generally accepted that the chemical effect of ultrasound in a chemical mixture Accepted: 17 June 2021 Published: 20 June 2021 results primarily from cavitation hot spots although it has been made clear that chemical transformations can also occur in ultrasonic fields without cavitation [1]. Due to locally
Publisher’s Note: MDPI stays neutral achieved extreme temperature and pressure conditions, an unusual chemical environ- with regard to jurisdictional claims in ment is often achieved using sonochemical method [2]. A growing body of research data published maps and institutional affil- shows that sonochemical synthesis of different materials, particularly in nanophases, is iations. useful and promising [3–5]. Using the ultrasonic technique, excellent emulsification and dispersion can be achieved [6,7]. Acoustic cavitation is also useful for efficient surface cleaning [8]. Sonochemical processing of semiconductor surfaces has been studied rather more recently [9–15]. Near-surface distribution of excess charge carriers can have very different properties in semiconductors and semiconductor nanostructures, depending on Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. morphology, temperature, and surface properties such as dangling bond defects of differ- This article is an open access article ent configurations, the presence of a suboxide interlayer, and bonds between adsorbates distributed under the terms and and surface atoms, etc. The sonochemical processing technique can therefore provide conditions of the Creative Commons a unique tool to modify the electronic properties following the response to the bubble Attribution (CC BY) license (https:// collapse and breaking the chemical bonds on the surface in a variety of materials. This is creativecommons.org/licenses/by/ particularly helpful for the processing of silicon whose surface electronic properties can 4.0/). very intriguingly be influenced by the chemical preparation [16].
Molecules 2021, 26, 3756. https://doi.org/10.3390/molecules26123756 https://www.mdpi.com/journal/molecules Molecules 2021, 26, 3756 2 of 16
The operating frequency range of sonochemical apparatus is typically up to dozens of kHz. A key assumption of choosing the frequency is that the size of the cavitation bubble is inversely related to the ultrasonic frequency. Therefore, because the bubble size drops with increasing the ultrasonic frequency and the bubble implosions become less violent, the energy released by each imploding cavitation bubble decreases with the ultrasonic frequency. However, the number of the imploding events increases due to the increased number of sound waves passing through the liquid at a higher frequency [17]. The computed bubble radius ranges from 0.1 to 100 µm at 20 kHz and from 0.1 to 3 µm at 1 MHz [18]. Above ≈1 MHz, the bubble disintegrates into smaller bubbles in a few acoustic cycles, while in the dozens and hundreds kHz frequency range, the bubble shape remains stable. The amount of water vapor in the collapsing bubble decreases with increasing the frequency. This behavior affects the multibubble sonoluminescence response, so that it originates mainly from the plasma emission at about 1 MHz while the involvement of OH radicals in the emissions was observed at 20 kHz [18]. Ultrasonic treatment at tuned frequencies has gained a great deal of interest in bio- medical research, e.g., being used to inactivate bacteria cells through cavitation [19]. In- terestingly relevant to the scope of this work, the inactivation effects, which are different at a low frequency of 20 kHz and higher frequencies of 580 and 1146 kHz, arise from the collapse of acoustic cavitation bubbles that generates both physical and chemical effects, as was reported previously [20,21]. Shock waves and shear forces produced by the bubble collapse and acoustic streaming are obviously among the physical effects. Free radicals, − − + HO ,O2 , and H , as well as other kinds of oxidants originate from the decomposition of water vapor within the collapsing bubble. If volatile solutes are dissolved in the sonicated mixture, they enter the bubbles by evaporation and are consequently dissociated as well. The resulting chemical products can then diffuse outside the bubble, dissolving in the surrounding liquid and producing a variety of chemical reactions. However, ultrasonic frequency is only one of the most important characteristic quanti- ties describing the chemical effect of ultrasound [22]. Thus, varying sonochemical activity with stirring is a very important feature of the sonochemical process [23]. Reactions occur- ring with a high reaction rate are randomly distributed inside a reaction vessel, so that the stirring time would be shorter than the reaction time in order to complete the reaction in sonochemical instruments [24]. Moreover, the frequency of irrigation is crucial for chemical processes utilizing sono- chemical devices [22]. In particular, passive ultrasonic irrigation implies the acoustic energy transmission aiming to activate the irrigant during the root canal treatment in dentistry [25]. Most generally, the pressure propagation through a liquid medium produces negative acoustic pressure, thus overcoming the tensile strength in the liquid medium and form- ing cavitation bubbles. In endodontics, this pressure change arises in the confined root canal [26]. Here, silicon wafers have been sonicated in tetrahydrofuran solutions. The surface properties were investigated by surface photovoltage (SPV) measurements. The results are reported and discussed. Tetrahydrofuran was used as a hydrogen promoter which, in turn, can induce the reconstruction of active centers at the Si surface that were associated with an appreciable modification of the charge carriers trapping and surface recombination dynamics. The influence of the ultrasonic frequency in the range from dozens to hundreds of kHz on the kinetic processes of photogenerated electrons and holes at silicon surfaces was demonstrated.
2. Materials and Methods Different ultrasonic experimental arrangements have been utilized to achieve sono- chemical treatments [3]. Most laboratory setups use either some ultrasonic bath or commer- cial ultrasonic horn systems. In this work, a low frequency sonication at about 28 kHz was performed in a standing wave ultrasonic bath using a Langevin transducer, as described elsewhere [27]. In developing a higher-frequency setup, it is important to keep in mind Molecules 2021, 26, x FOR PEER REVIEW 3 of 16
Molecules 2021, 26, 3756 commercial ultrasonic horn systems. In this work, a low frequency sonication at about3 of 2 168 kHz was performed in a standing wave ultrasonic bath using a Langevin transducer, as described elsewhere [27]. In developing a higher-frequency setup, it is important to keep in mind that the time for bubble creation may be longer in this case. Thus, at 20 kHz, the rarefactionthat the time cycle for bubbleis 25 μs creation with peak may pressure be longer amplitude in this case. achieved Thus, at in 20 12.5 kHz, μs, the whereas rarefaction the appropriatecycle is 25 µ cycles with is peak only pressure0.025 μs at amplitude the frequency achieved of 20 in MHz 12.5 [µ28s,]. whereas Consequently, the appropriate one can concludecycle is only that 0.025 the bubbleµs at the production frequency likelihood of 20 MHz drops [28]. Consequently,with the frequency, one can so conclude that greater that acousticthe bubble powers production should likelihood be used in drops order with to theovercome frequency, the socohes thative greater forces acoustic in a chemical powers mixtureshould beover used shorter in order periods to overcome of the rarefaction the cohesive cycles. forces For in aexample, chemical increasing mixture over the shorter oper- atingperiods frequency of the rarefaction from 10 to cycles. 400 kHz For requires example, an increasing order of themagnitude operating enhancement frequency from in ul- 10 trasonicto 400 kHz power requires to maintain an order cavitation of magnitude in water enhancement [29]. The solid in ultrasonic horns are power most tocommonly maintain employedcavitation in watersonochemistry, [29]. The solidmaterial horns forming, are most processing commonly, and employed medical in applications sonochemistry, to delivermaterial h forming,igh-intensity processing, ultrasonic and energy. medical Variousapplications device to deliver designs high-intensity and design ultrasonic formulas haveenergy. been Various widely device considered designs and [29, design30]. An formulas interesting have multi been widely-domain considered high-frequency [29,30]. An interesting multi-domain high-frequency LiNbO3 plate design has been proposed LiNbO3 plate design has been proposed towards miniaturization of the resonator [31]. towards miniaturization of the resonator [31]. Here, a piezoceramic transducer operating at about 400 kHz was designed to max- Here, a piezoceramic transducer operating at about 400 kHz was designed to maximize imize the acoustic power stored in the ultrasonic bath and thus to improve the energy the acoustic power stored in the ultrasonic bath and thus to improve the energy efficiency efficiency of the sonochemical processing process. of the sonochemical processing process.
2.1.2.1. Designing Designing Ultrasonic Ultrasonic Transducer Transducer AA processed processed silicon wafer sample isis placedplaced atat thethe bottom bottom of of a a flask flask filled filled with with a liquida liq- uidreactant reactant mixture, mixture, as shownas shown in Figurein Figure1. The1. The piezoceramic piezoceramic transducer, transducer, placed placed between between a afront front metal metal electrode electrode and and back back metal metal plate, plate is, externallyis externally glued glued to theto the bottom bottom surface surface of theof theflask flask and and generates generates an ultrasonican ultrasonic wave wave inside inside the the mixture. mixture.
Figure 1. Schematics of the setup used for sonication at about 400 kHz. Figure 1. Schematics of the setup used for sonication at about 400 kHz. We considered a piezoceramic disk with the thickness of L and poled in the z− direction,We consider as showned a inpiezoceramic Figure2a. The disk left with and the right thickness sides of the퐿 and plate poled were in metalized.the 푧 −di- rectionIn the, general as shown case, in Figure the vibration 2a. The of lef thet and piezoelectric right sides bodyof the is plate described were metalized. by the following In the generalequations case, [32 the]: vibration of the piezoelectric body is described by the following equa- 2 tions [32]: ∂ u ∂Tij ρ i = , (1) 22 ∂t휕 푢푖 ∂휕x푇j푖푗 휌 2 = , (1) 휕푡 휕푥푗 ∂Di 휕퐷푖 = 0, (2) ∂xi = 0, (2) 휕푥푖 ∂u ∂ϕ = E 퐸 휕k푢푘+ 휕휑 Tij푇푖푗 =cijkl푐푖푗푘푙 +e푒mij푚푖푗 ,, ( (3)3) ∂x휕l푥푙 ∂휕x푥m푚 ∂u휕푢푘 푆 ∂ϕ휕휑 퐷푖 = 푒푖푘푙 k − 휀S푖푗 , (4) Di = eikl 휕푥푙− εij 휕푥푗 , (4) ∂xl ∂xj
where xj is the coordinate (j = 1, 2, 3), ui is the displacement vector component (i =1, 2, 3), t is time, ρ is the mass density, Tij is the stress tensor, Di is the electric displacement vector, E cijkl are the elastic moduli of the medium under constant electric field, emij are its piezoelec- Molecules 2021, 26, x FOR PEER REVIEW 4 of 16
Molecules 2021, 26, 3756 where 푥푗 is the coordinate (푗 = 1, 2, 3), 푢푖 is the displacement vector component4 of (푖 16= 1, 2, 3), 푡 is time, 휌 is the mass density, 푇푖푗 is the stress tensor, 퐷푖 is the electric dis- 퐸 placement vector, 푐푖푗푘푙 are the elastic moduli of the medium under constant electric field, 푒 are its piezoelectric coefficients, 휀푆 is the permittivity tensor under constant strain tric coefficients,푚푖푗 εS is the permittivity tensor푖푗 under constant strain S = 1 ∂ui + ∂uk , and 1 휕푢푖 휕ij푢푘 ik 2 ∂xk ∂xi 푆푖푘 = ( + ), and repeated indices are summed. repeated2 indices휕푥푘 are휕푥푖 summed.
2
Tai, uai 1 L Tbi, ubi x 4
z y 1 Vi Ii 2 3
V
(a) (b)
FigureFigure 2. (a 2.) Schematics(a) Schematics of the of the longitudinal longitudinal piezoelectric piezoelectric transducer: transducer: 1—piezoceramic 1—piezoceramic disk, disk 2—metal, 2—metal electrode electrode plates, plates, 푉 3—working3—working surface surface of the of transducer,the transducer, displacement displacement directions, directions,V—applied —applied rf voltage; rf voltage; (b) six-pole (b) six network-pole network model model (1) for the(1) for the transducer with an acoustic load (2). transducer with an acoustic load (2).
GivenGiven that that the the thickness thickness of the of transducerthe transducer was muchwas much smaller smaller than thethan diameter the diameter of the of disk,the we disk could, we considercould consider a one-dimensional a one-dimensional (1D) vibration (1D) vibration body. body. Assuming Assuming the wave the wave is a harmonicis a harmonic plane plane wave wave traveling traveling in the in zthe− direction,푧 −direction, the displacementthe displacement components components are are defineddefined by: by: j( t−kz) u = u e ω 푗(휔푡,−푘푧) (5) l 푢푙0=l 푢0푙푒 , (5) wherewherej is the푗 is imaginarythe imaginary unit, unit,k is the푘 is wave the wave number, number and ω, andis the 휔 angularis the angular frequency frequency of the of 0 00 wave.the Usingwave.complex Using complex wave numbers wave numbersk = k − jk푘 =allowed푘′ − 푗푘 us′′ allow to takeed into us accountto take into the losses account in thethe transducer.losses in the transducer. SubstitutingSubstituting Equation Equation (5) (5) with withl =푙 =1, 1, 2, 2, 3 3into into EquationsEquations (1)–(4)(1)–(4) provided a a few few so- solutions for one longitudinal and two transverse waves. In our particular case, only the lutions for one longitudinal and two transversep waves. In our particular case, only the longitudinal wave with the velocity of V = c /ρ was relevant to the transducer shown longitudinal wave with the velocityl of 푉33= √푐 /휌 was relevant to the transducer in Figure2a. In this case, there would be two longitudinal푙 33 waves counterpropagating in shown in Figure 2a. In this case, there would be two longitudinal waves counterpropa- the z and −z directions. Omitting the exponential term ejωt and taking into account that gating in the 푧 and −푧 directions. Omitting the exponential term 푒푗휔푡 and taking into only the E3(z) component of the electric field strength is not equal to zero, one equates: account that only the 퐸3(푧) component of the electric field strength is not equal to zero, one equates: −jkz jkz u1 = a1e + b1e , (6) −푗푘푧 푗푘푧 푢1 = 푎Z1푒 + 푏1푒 , (6) ϕ = − E3dz (7) 휑 = − ∫ 퐸3푑푧 (7) The electrical current through the transducer is given by I = jωAD3, with A being the cross-section of the transducer. The electrical current through the transducer is given by 퐼 = 푗휔퐴퐷3, with 퐴 being theDividing cross-section the transducer of the transducer. into a stack of n thin layers, the displacement in the i-th layer takes theDividing form: the transducer into a stack of 푛 thin layers, the displacement in the i-th −jk(z−zi) jk(z−zi) layer takes the form: u1 = a1e + b1e , (8)
which results in the stress in the i-th layer: −푗푘(푧−푧푖) 푗푘(푧−푧푖) 푢1 = 푎1푒 + 푏1푒 , (8) −jki(z−zi) jki(z−zi) which results in theT stressi = ci jkini the−a ie-th layer: + bie − eiE3, (9)
−푗푘푖(푧−푧푖) 푗푘푖(푧−푧푖) 푇푖 = 푐푖푗푘푖(−푎푖푒 + 푏푖푒 ) − 푒푖퐸3, (9)
Molecules 2021, 26, 3756 5 of 16
E 2 S where ci = c33 + e33/ε33 and ei = e33 for the i-th layer. The voltage drop across the i-th layer is then given by:
Z zi+1 e I L i −jki Li jki Li i i Vi = − E3dz = ai e − 1 + bi e − 1 + j , (10) zi εi ωAεi
where εi and Ii are the permittivity and the current in the i-th layer, respectively, and Li = zi+1 − zi is the thickness of the i-th layer. One can then employ the impedance matrix approach to describe the transducer [32,33]. For this, the boundary conditions for the i-th layer take the form:
uai = ui−1| = ui| , ubi = ui| = ui+1| , z=zi z=zi z=zi+1 z=zi+1 (11) T = T | = T | , T = T | = T | ai i−1 z=zi i z=zi bi i z=zi+1 i+1 z=zi+1
Using Equations (10) and (11), one writes the following system of three linear equations: − ciki ciki jei Tai tan ki Li sin ki Li ωAεi uai ciki ciki jei Tbi = − · ubi (12) sin ki Li tan ki Li ωAεi Vi − ei ei jLi Ii εi εi ωAεi
S where ei = e33 and εi = ε33. The square 3 × 3 matrix was a three-port impedance matrix. Since each of the three ports had two transmission lines, a six-pole network was formed, as shown in Figure2b. The network had two acoustic ports, which are marked as Tai, uai and Tbi, ubi in Figure2b, and one electrical port (Vi). In the case of non-piezoelectric vibrating disks, the electrical port was absent in the network, so one would consider only two acoustic ports shown in Figure2b. Hence, a solid layer of finite thickness is a two-port element forming a four-pole network. With increasing layer thickness to infinity, one would attain a two-pole network or a one-port element, which acts as the absorber of ultrasonic waves. At the absorber edge, the amplitude of the forward wave traveling from infinity to the port was equal to zero such that the stress and displacement could be related through the acoustic impedance Z as Tb = Zbub. If the six-pole network p is acoustically loaded by one-port elements with the impedances Za and Zb at the left- and right-hand sides, respectively, the resistance of the appropriately loaded transducer can be found as:
p3,1(Zb p1,3 + p1,2 p2,3 − p1,3 p2,2) + p3,2(Za p1,2 + p1,3 p2,1 − p1,1 p2,3) R = p3,3 + (13) ZaZb − Za p2,2 − Zb p1,1 + p1,1 p2,2 − p1,2 p2,2
The four-pole network q connected to a one-port element with the impedance Zb at the right-hand side, as shown in Figure2b, becomes a two-pole network with:
Zbq1,1 − q1,1q2,2 + q1,2q2,1 Za = (14) Zb − q2,2
Instead, the four-pole network q connected to a one-port element with the impedance Za at the left-hand side has:
Zaq2,2 − q1,1q2,2 + q1,2q2,1 Zb = (15) Za − q1,1
Therefore, using Equations (14) and (15), the transducer with any number of loading layers can be described by the two-pole network with the resistance given by Equation (13). Our computation result is shown by curve 1 in Figure3. Molecules 2021, 26, 3756 6 of 16 Molecules 2021, 26, x FOR PEER REVIEW 6 of 16
Figure 3. Calculated (1) and measured (2) impedance magnitude response for piezoelectric transducer Figure 3. Calculated (1) and measured (2) impedance magnitude response for piezoelectric trans- shown in Figure2a. ducer shown in Figure 2a. To experimentally realize the ultrasonic reactor setup shown in Figure1, which oper- To experimentally realize the ultrasonic reactor setup shown in Figure 1, which op- ates in the frequency range from about 300 to 500 kHz, we used a round-bottom glass flask erates in the frequency range from about 300 to 500 kHz, we used a round-bottom glass with a diameter of about 40 mm and a bottom thickness of 3.1 mm. Front brass electrode flask with a diameter of about 40 mm and a bottom thickness of 3.1 mm. Front brass was about 0.1 mm, whereas the back metal plate and electrode (see Figure1) were round electrodedisks of 40was mm about in diameter 0.1 mm, and whereas 2 mm the in thickness. back metal plate and electrode (see Figure 1) were roundThe measured disks of frequency40 mm in diameter dependence and of 2 themm impedance, in thickness. which is shown by curve 2 in FigureThe3 ,measured demonstrated frequency an excellent dependence agreement of the with impedance, the computed which impedanceis shown by response curve 2 in(curve Figure 1). 3, It was demonstrate reliably identifiedd an excellent that the agreement resonance with at about the 400 computed kHz marked impedance by arrow re- in sponseFigure (3curve was due1). It to was a coupled reliably oscillator, identified which that the represented resonance theat about lowest-vibration 400 kHz marked mode by in arrowan acoustic in Figure system 3 wa composeds due to a coupled of the piezoceramic, oscillator, which backing, represent fronted metal the lowest layers,- andvibration water modeload (seein an Figure acoustic1). Scanning system composed the driving of frequency, the piezoceramic we observed, backing, intense front acoustic metal cavitation layers, andin awater liquid load poured (see into Figure the flask 1). Scanning just at the the frequency driving of frequency, about 400 kHz,we observe thus verifyingd intense the acousticessential cavitation correctness in a of liquid the transducer poured into fabrication the flask andjust at quantitation the frequency approach. of about 400 kHz, thus verifyingTo compare the sonochemical essential correctness efficiency of at the the frequenciestransducer fabricationof about 28 and 400 quantitation kHz used approach.here, standard calorimetric measurements [34] were performed. For this purpose, 20 mL of waterTo wascompare sonicated sonochemical at 28 and efficiency 400 kHz. Inat the both frequencies cases, the ultrasonicof about 28 power and 400W kHzdissipated used here,into standard the water calorimetric was calculated measurements as [34]: [34] were performed. For this purpose, 20 mL of water was sonicated at 28 and 400 kHz. In both cases, the ultrasonic power 푊 dissi- ∆T pated into the water was calculated as W[34=]: mCp, (16) ∆t ∆푇 푊 = 푚퐶푝, (16) where ∆T/∆t is the temporal rate of the temperature∆푡 (T) rise, m is the mass of water, and wCherep = 4.2∆푇/ J/g∆푡 Kis isthe the temporal specific heatrate of waterthe temperature at constant ( pressure.푇) rise, 푚 Theis the initial mass temperature of water, andrise 퐶 was푝 = measured4.2 J/g K is at the room specific temperature heat of bywater using at constant a DS18B20 pressure. digital thermometer,The initial temper- which aturewas immersedrise was measured in the water at room and wastemperature held at its by half using height. a DS18B20 The resulting digital acousticthermometer, power whichwas set was to immersed about 20 W, in whichthe water corresponded and was held to acousticat its half power height. densities The resulting of ≈1 W/mLacoustic at powerboth the was frequencies set to about used. 20 W, which corresponded to acoustic power densities of ≈1 W/mLHowever, at both the this frequencies value gives used. the net ultrasonic power dissipated in the liquid [35]. So, anotherHowever, test of this this value type, gives estimating the net the ultrasonic sonochemical power effectiveness, dissipated in was the made liquid using [35]. aSo, KI − aoxidationnother test dosimetry of this type, technique. estimating The the iodide sonochemical ions (I ) ineffectiveness, the aqueous was solution made ofusing KI can a KI be − oxidationtransformed dosimetry into iodine techniqu moleculese. The ( Iiodide2) under ions sonication. (I−) in the If aqueous excess I solutionions are presentof KI can in be the − I − transformedsolution, 2 willinto furtheriodine molecules react with ( themI2) under thus sonication. forming the If triiodideexcess I ions ions are I3 presentas [36] in the − solution, I2 will further react with them thus −forming− the triiodide ions I3 as [36] I2 + I ↔ I3 (17) 퐼 + 퐼− ↔ 퐼− (17) 2 − 3 The formation of the yellow complex I3 in the colored solution was controlled by − theThe characteristic formation opticalof the yellow absorption complex peaks I3 in at the about colored 290 solution and 350 was nm controlled [37,38]. In by these the characteristic optical absorption peaks at about 290 and 350 nm [37,38]. In these experi-
Molecules 2021, 26, x FOR PEER REVIEW 7 of 16 Molecules 2021, 26, 3756 7 of 16
ments, concentration of KI in water was 0.1 mol/L. Exposure to ultrasound changed ini- experiments,tially clear solution concentration to yellow of KI colored. in water The was absorption 0.1 mol/L. spectra Exposure were to ultrasoundmeasured using changed an initiallyEvolution clear 600 solution Spectrophotometer. to yellow colored. The spectral The absorption changes spectra were found were measuredto be quite using similar an Evolutionfor sonication 600 Spectrophotometer.at 28 and 400 kHz but The the spectral peak absorption changes were in the found 350 nm to be band quite was similar as much for sonicationas several times at 28 andgreater 400 at kHz 400 but kHz the than peak that absorption at 28 kHz. in the 350 nm band was as much as severalWithin times the greater frame at of 400 fluid kHz mechanics, than that atcavitation 28 kHz. can be described as being due to the impulsiveWithin formation the frame of of cavities fluid mechanics,caused by tensile cavitation forces can in behigh described-speed flows as being or flow due gra- to thedients impulsive arisen in formation a liquid. The of cavities flow pattern caused can by be tensile related forces to the in high-speedcharacteristic flows distribution or flow gradientsof nodes and arisen antinodes in a liquid. along The the flow length pattern of canthe beoscillating related to liquid. the characteristic Consequently, distribution the dis- oftribution nodes of and sound antinodes pressure along at the the transducer length of fundamental the oscillating frequency liquid. Consequently,plays an important the distributionrole for various of sound sonochemical pressure at reaction the transducers [39,40 fundamental]. The measured frequency sound plays pressure an important is quite roleaccurate for various and comparable sonochemical to that reactions simulated [39 numerically,40]. The measured [40,41]. However, sound pressure spatial isareas quite of accuratehigh pressures and comparable do not necessarily to that simulated correspond numerically to the ones [40 with,41]. high However, reaction spatial rates areasbecause of highthe reaction pressures is mediated do not necessarily by the bubbles. correspond In the to ultrasound the ones with process, high reactionvarious ratesmicrobubbles because the reaction is mediated by the bubbles. In the ultrasound process, various microbubbles inside of solvent vapor are formed, and that produces acoustic energy with radial motion inside of solvent vapor are formed, and that produces acoustic energy with radial motion through the reaction medium. Moreover, various microbubbles, from 4 to 300 μm in through the reaction medium. Moreover, various microbubbles, from 4 to 300 µm in diameter, can be formed inside the solvent vapor, thus producing acoustic energy flows diameter, can be formed inside the solvent vapor, thus producing acoustic energy flows in in the radial direction [42,43]. When the resonance frequency of the microbubbles exceeds the radial direction [42,43]. When the resonance frequency of the microbubbles exceeds that of the ultrasonic field, they collapse, which triggers the biochemical or thermo- that of the ultrasonic field, they collapse, which triggers the biochemical or thermochemical chemical reactions [22]. We did not address this issue here, restricting ourselves to de- reactions [22]. We did not address this issue here, restricting ourselves to describing the scribing the pressure distributions at the transducer resonance frequencies. pressure distributions at the transducer resonance frequencies. The results of our numerical simulations are shown in Figure 4. Following extensive The results of our numerical simulations are shown in Figure4. Following extensive discussions, one may expect that the bubbles are repelled from the pressure antinodes discussions, one may expect that the bubbles are repelled from the pressure antinodes due due to Bjerknes force and settled at locations between the pressure antinodes and nodes to Bjerknes force and settled at locations between the pressure antinodes and nodes [40]. In our[40]. experiments, In our experiments, the wafer the sample wafer sample was placed was justplace ind thejust region in the betweenregion between the antinodes the an- andtinodes nodes and when nodes using when the using lower the frequency lower frequency sonication. sonication. At about 400At about kHz, it400 resided kHz, nearit re- thesided bottom near the of the bottom flask of shown the flask in Figure shown4. in Figure 4.
(a) (b)
FigureFigure 4.4. SimulatedSimulated distributiondistribution ofof soundsound pressurepressure atat thethe drivingdriving frequencyfrequency ofof 2828 ((aa)) andand 400400 kHzkHz ((bb).). TheThe flaskflask isis 2020 mmmm inin diameter,diameter, 3232 mmmm inin height,height, withwith thethe wallwall thicknessthickness ofof 22 mm. mm. TheThe numericalnumerical simulationssimulations werewere performedperformed usingusing thethe COMSOLCOMSOL software.software.
2.2.2.2. SurfaceSurface ProcessingProcessing andand PhotovoltagePhotovoltage MeasuringMeasuring TheThe measurementsmeasurements werewere performedperformed onon p-type,p-type, 1–201–20 ΩΩ cm,cm, Cz-Si(111)Cz-Si(111) waferswafers withwith aa thicknessthickness ofof 330330 µμm.m. DistilledDistilled waterwater waswas usedused asas aa solvent.solvent. AllAll reagentsreagents werewere analyticalanalytical gradegrade andand usedused asas purchasedpurchased withoutwithout furtherfurther purificationpurification andand treatment.treatment. AllAll thethe weightweight ◦ measurementsmeasurements were were done done at at about about 20 20C usingC using an analytical an analytical balance. balance. The freshlyThe freshly prepared pre- mixturespared mixtures were stirred were forstirred about for 1 about min and 1 min then and ultra-sonicated then ultra-sonicated for 15 min forto 15 ensure min to solvent ensure homogeneity.solvent homogeneity. Reactant Reactant solutions solutions were not were intentionally not intentionally air-saturated air-saturated before their before use their and useduse and within used 30 within min after 30 min the preparation.after the preparation The sample. The temperature sample temperature was kept was constant kept atconstant about at 310 about K using 310 K a using fan heatsink. a fan heatsink. The DS18B20 The DS18B20 temperature temperature probe probe was w placedas placed in
MoleculesMolecules 20212021,, 2626,, x 3756 FOR PEER REVIEW 8 8of of 16 16
inthe the irradiated irradiated solution solution to to control control the the temperature temperature inside inside it. it. The The temperature temperature monitoring monitor- ingcurves curves for for the the solution solution irradiated irradiated at at 28 28 and and 400 400 kHz kHz are are shown shown in in FigureFigure5 .5. BeforeBefore thethe measurements,measurements, the temperature ofof thethe liquid liquid inside inside the the flask flask was wa thes the ambient ambient temperature. tempera- ture.After After the sonication the sonication was turnedwas turned on, the on, temperature the temperature increase increase∆T was ∆푇 monitored, was monitored, as shown as shownby curves by 1curves and 2 1 in and Figure 2 in5 .Figure At both 5. frequencies,At both frequencies, the solution the temperature solution temperature was stabilized was stabilizedat the larger at the value larger of ∆ valueT ≈ 17–22 of ∆푇 K≈about 17–22 6 K min about after 6 min starting after the starting sonication. the sonication. After the Aftertemperature the temperature stabilization, stabilization, the sample the was sample placed was into theplaced solution, into the and solution the wafer, a surfacend the waferprocessing surface was processing performed. was performed.
Figure 5. Variation of solution temperature with time after the sonication at 28 (curve 1) and 400 (2) Figure 5. Variation of solution temperature with time after the sonication at 28 (curve 1) and 400 (2) kHzkHz is is turned turned on on.. Before the chemical and sonochemical processing, the Si wafers were cleaned by a Before the chemical and sonochemical processing, the Si wafers were cleaned by a cleaning procedure, which can include rinsing with water, methanol, acetone, methanol, cleaning procedure, which can include rinsing with water, methanol, acetone, methanol, and finally water [44,45]. Here, we used tetrahydrofuran instead of acetone. Following the and finally water [44,45]. Here, we used tetrahydrofuran instead of acetone. Following above cleaning treatments, sonication was performed in tetrahydrofuran (THF) C H O, the above cleaning treatments, sonication was performed in tetrahydrofuran (THF)4 8 methanol CH3OH and in a solvent mixture containing THF and methanol (THF/methanol C4H8O, methanol CH3OH and in a solvent mixture containing THF and methanol volume ratio was equal to 3/1 and 1/1) for 15 min. (THF/methanol volume ratio was equal to 3/1 and 1/1) for 15 min. The surface photovoltage transients were measured in the contactless capacitor ar- The surface photovoltage transients were measured in the contactless capacitor ar- rangement using the pulsed modulated light and a 100-µm glass dielectric spacer as rangementdescribed elsewhereusing the pulsed [46]. The modulated decay of thelight SPV and was a 100 taken-μm afterglass the dielectric light emitting spacer as diode de- scribed(center wavelengthelsewhere [4 at6]. 405 The nm) decay was switchedof the SPV off. was taken after the light emitting diode (center wavelength at 405 nm) was switched off. 3. Results and Discussion 3. ResultsIllumination and Discussion of the silicon surface induces excess free carriers, which, being sepa- ratedIllumination in the near-surface of the silicon region, surface cause induces an occurrence excess offree the carriers, surface which, photovoltage. being sepa- The ratedsurface-sensitive in the near- SPVsurface can region, therefore cause be used an occurrence to investigate of the influence surface photovoltage. of important wet-The surfacechemical-sensitive treatments SPV on can the thereforeelectronic besurface used properties to investigate as well the as influencethe hydrogen of important and oxide wetcoverage-chemical [47]. treatments on the electronic surface properties as well as the hydrogen and oxideThe coverage surface [4 passivation7]. implies the achievement of a rather long effective minority car- rier lifetimeThe surfaceτe ranging passivation on a time implies scale the of milliseconds,achievement whichof a rather is affected long byeffective the contribution minority carrierfrom both lifetime the surfaces 휏푒 ranging and on the a bulk timeτ bscaleas [48 of]: milliseconds, which is affected by the con- tribution from both the surfaces and the bulk 휏푏 as [48]: 1 1 2υe f f 1= 1+ 2휐푒푓푓 , (18) τe =τ + W , (18) 휏푒 b휏푏 푊
wherewhere 휐υ푒푓푓e f f isis the effective effective surface surface recombination recombination velocity velocity and and 푊W isis the the wafer wafer width. width. It is It thereforeis therefore clear clear that that the the surface surface recombination recombination proc processes,esses, which which are are accounted accounted for for in in EquationEquation (18)(18) by by introducing introducingυ e휐 f f푒푓푓, point, point to a to fundamental a fundamental physical physical limitation limitation in increasing in in-
Molecules 2021, 26, x FOR PEER REVIEW 9 of 16
creasing 휏푒. The problem can be circumvented by employing surface processing, e.g., sonication, which affects the carrier recombination velocities at the surface. Figure 6a shows the variation of the SPV decays taken before (curves 1, 3, and 5) and after its chemical (curve 2) and sonochemical (curves 4 and 6) treatments in tetrahydro- furan. Exposure to lower-frequency sonication at about 28 kHz yields decay 4, while higher-frequency treatments at about 400 kHz result in decay 6 in Figure 6a. Figure 6b displays appropriate results obtained in a mixture of tetrahydrofuran and methanol. In order to check the post-treated evolution of the decays, the SPV transients were collected one day after appropriate chemical and sonochemical treatments had been made. Typical Molecules 2021, 26, 3756 9 of 16 results obtained for one sample set are shown by curves 7–9 in Figure 6b, illustrating that allowing the treated sample to stay at rest was not able to remove the procession-induced changes sufficiently. τe. TheThe problem most prominent can be circumvented effect in Figure by employing6a is the remarkable surface processing, SPV enhancement e.g., sonication, due to whichincreasing affects sonication the carrier frequency. recombination More velocities conveniently, at the surface. this can be seen in Figure 7a, showingFigure the6a relative shows thechange variation in the ofSPV the amplitudes SPV decays 푈 taken0 taken before at time (curves 푡 = 1,0, 3,just and when 5) and the afterlight its is chemicalturned off, (curve due 2) to and chemical sonochemical treatment (curves and 4 and sonication 6) treatments in THF in tetrahydrofuran.. It is seen that Exposurechemical treatment to lower-frequency in tetrahydrofuran sonication quenches at about the 28 SPV kHz signal yields (at decay most 4, a while60% dec higher-rease frequencyin Figure 7 treatmentsa). A noticeably at about smaller 400 kHzdecrease, result approximately in decay 6 in Figure 27%, 6isa. achieved Figure6b sonochem- displays appropriateically with a resultsfrequency obtained of 28 kHz. in a mixtureIn marked of contrast, tetrahydrofuran the SPV andresponse methanol. is increased In order up toto check≈80% due the to post-treated sonication evolutionat 400 kHz of. the decays, the SPV transients were collected one day afterThe time appropriate-dependent chemical SPV decays and sonochemical 푈(푡) can be wr treatmentsitten as a hadproduct been of made. the amplitude Typical resultsvalue of obtained the SPV for signal one sample푈0 and set the are shapefunction shown by curves 휂 (푡) 7–9, 푈 in(푡) Figure= 푈0휂6 (b,푡) illustrating. Depending that on allowingthe separation the treated and recombination sample to stay processes at rest was associated not able to with remove photoexcited the procession-induced electrons and changesholes, 휂( sufficiently.푡) can be approximated, e.g., by the single- or multiexponential form.
(a) (b)
FigureFigure 6.6. Time-dependentTime-dependent SPVSPV ofof SiSi samplessamples beforebefore (curves(curves 1,1, 3,3, andand 5)5) andand afterafter chemicalchemical (curves(curves 2)2) andand sonochemicalsonochemical treatmentstreatments atat aboutabout 2828 kHzkHz (curves(curves 4)4) andand 400400 kHz kHz (curves (curves 6): 6): ( a(a)) in in tetrahydrofuran; tetrahydrofuran; ( b(b)) in in tetrahydrofuran/methan tetrahydrofuran/methan olol mixture with a volume ratio of 3/1. Green curves 7–9 exemplify the SPV evolution one day after appropriate chemical and sonochemical treatments were made.
The most prominent effect in Figure6a is the remarkable SPV enhancement due to increasing sonication frequency. More conveniently, this can be seen in Figure7a, showing the relative change in the SPV amplitudes U0 taken at time t = 0, just when the light is turned off, due to chemical treatment and sonication in THF. It is seen that chemical treatment in tetrahydrofuran quenches the SPV signal (at most a 60% decrease in Figure7a ). A noticeably smaller decrease, approximately 27%, is achieved sonochemically with a frequency of 28 kHz. In marked contrast, the SPV response is increased up to ≈80% due to sonication at 400 kHz. Molecules 2021, 26, x FOR PEER REVIEW 10 of 16
Molecules 2021, 26, 3756 10 of 16 mixture with a volume ratio of 3/1. Green curves 7–9 exemplify the SPV evolution one day after appropriate chemical and sonochemical treatments were made.
(a) (b) 휏 FigureFigure 7.7.Relative Relative changechange inin thethe SPVSPV amplitudesamplitudes ((aa)) andand decaydecay timestimes τ00( b(b)) due due to to exposure exposure of of Si Si to to tetrahydrofuran tetrahydrofuran and and sonicationsonication in in tetrahydrofuran tetrahydrofuran at aboutat about 28 and 28 and 400 kHz 400 kHz (red rectangles).(red rectangles) Black. rectanglesBlack rectangles are measured are measur whened using when a mixture using a mixture of tetrahydrofuran and methanol. The data in (a) are obtained from Figure 4 at time moment 푡 = 0 and that in of tetrahydrofuran and methanol. The data in (a) are obtained from Figure4 at time moment t = 0 and that in (b) by fitting (b) by fitting the decays in Figure 6a to Equation (19). the decays in Figure6a to Equation (19).
TheFollowing time-dependent the analysis SPV given decays elsewhereU(t) can [4 be9,50 written], a stretched as a product-exponent of the decay amplitude model was used here, so that the SPV signal 푈(푡) is described by: value of the SPV signal U0 and the shapefunction η (t), U(t) = U0η (t). Depending on the 훽 separation and recombination processes associated−(푡/ with휏0) photoexcited electrons and holes, 푈(푡) = 푈0푒 , (19) η(t) can be approximated, e.g., by the single- or multiexponential form. whereFollowing 휏0 is the the characteristic analysis given stretched elsewhere-exponent [49,50 ], decay a stretched-exponent time and 훽 is the decay dispersion model wasfactor, used which here, describes so that the the SPV spread signal ofU time(t) is constants. described Obviously, by: 훽 = 1 for a monoexpo- nential decay, whereas the values 0 < 훽 < 1 describe a variation from a monoexponen- β −(t/τ0) tial decay with smaller 훽 values correspondingU(t) = U0e to a ,broader distribution of decay times. (19) All the fitting results are summarized in Table 1. where τ0 is the characteristic stretched-exponent decay time and β is the dispersion factor, Table 1. The peak amplitudeswhich 푈0 describesof the SPV thesignals spread and fitting of time parameters constants. 휏0 Obviously,and 푏 for theβ dec= 1ays for shown a monoexponential in Figure 6. decay, whereas the values 0 < β < 1 describe a variation from a monoexponential decay Treatment Peak Amplitude (mV) 흉 (μs) 휷 with smaller β values corresponding to a broader distributionퟎ of decay times. All the fitting results are summarized in112.8 Table (curve1. 1) 6.0 0.294 44.4 (curve 2) 11.1 0.329 THF Table 1. The peak amplitudes111.5U0 of (curve the SPV 3) signals and fitting parameters9.0 τ0 and b for0.321 the decays Figure 4a shown in Figure6. 83.6 (curve 4) 5.1 0.288 41.9 (curve 5) 10.8 0.335 Treatment Peak Amplitude (mV) τ0 (µs) β 75.6 (curve 6) 8.4 0.296 112.8 (curve 1) 6.0 0.294 14.6 (curve 1) 6.4 0.302 44.4 (curve 2) 11.1 0.329 THF 15.4 111.5(curve (curve 2) 3)3.0 9.00.247 0.321 THF/Methanol (3/1) Figure4a 41.4 (curve83.6 (curve 3) 4)7.3 5.10.307 0.288 Figure 4b 74.6 (curve41.9 (curve 4) 5)5.3 10.80.275 0.335 20.0 (curve75.6 (curve 5) 6)10.3 8.40.309 0.296 34.4 (curve14.6 (curve 6) 1)5.7 6.40.274 0.302 15.4 (curve 2) 3.0 0.247 THF/Methanol (3/1) 41.4 (curve 3) 7.3 0.307 It Figureis seen4b in Table 1 that the74.6 factor (curve 4)훽 is not affected greatly 5.3 under different 0.275 treat- ments in tetrahydrofuran and is20.0 approximately (curve 5) equal to 0.3 for 10.3 the data shown 0.309 in Figure 6a. In contrast, the decay time 34.4휏0 varies (curve 6)rather significantly upon 5.7 different kinds 0.274 of the treatments. After chemical treatment in tetrahydrofuran, the time constant sharply in- creasesIt is by seen about in Table 80%1 fr thatom theits initial factor valueβ is not. In affectedturn, the greatly sonication under in different THF leads treatments to short- inening tetrahydrofuran of the decay, andmore is pronounced approximately for equalthe sonochemical to 0.3 for the processing data shown at inabout Figure 28 6kHza. In. contrast, the decay time τ0 varies rather significantly upon different kinds of the treatments. After chemical treatment in tetrahydrofuran, the time constant sharply increases by about
Molecules 2021, 26, 3756 11 of 16
80% from its initial value. In turn, the sonication in THF leads to shortening of the decay, more pronounced for the sonochemical processing at about 28 kHz. The likely mechanism that has come to describe the previous observations with chloroform CHCl3 and dichloromethane CH2Cl2 relies on the fact that these solutions act as sources of carbon, which can saturate the dangling bonds revealed on the surface of Si, thus passivating the surface [15,51]. In this scenario, using CHCl3 or CH2Cl2 yields the Si–H and C–Cl bonds, which react, forming C–H species. This, in turn, resembles the chlorination/alkylation process that forms Si−alkyl, converting Si–H into Si–CnH2n+1 (n ≥ 1). The alkyl chains on Si surfaces are known to provide low surface recombination velocities [52], thus featuring effective Si surface passivation [44]. Too little is still known about the nature and mechanisms of formation of the ini- tial radicals and molecules produced at thermal decomposition of the tetrahydrofuran ring [53–55]. Meanwhile, the overall pyrolysis of THF can be summarized by the following major participating reactions [54]:
THF → CH2 = CH2 + (CH2)2 − O, (20)
(CH2)2 − O → HCO + CH3, (21) HCO → H + CO, (22)
THF → CH3 − CH = CH2 + CH2O, (23)
CH3 − CH = CH2 → CH2 = C = CH2 → CH3 − C ≡ CH, (24)
CH3 − CH = CH2 + CH3 → CH3 − CH2 − CH = CH2 + H (25) Here, Equation (24) gives a schematic reaction presentation. It is seen in Equation (22), Equation (24), and Equation (25) that hydrogen and carbon (as well as carbon monoxide) can be among the by-products. Therefore, the atomic hydrogen is able to effectively passivate dangling bonds at the Si surface, suppressing the surface state density and surface recombination velocities [56], and hence enhancing the SPV response. Furthermore, carbon atoms at the surface create Si−C bonds and dangling carbon bonds, which are then saturated by H atoms. Methanol can be used to dilute the solvent and adjust the molar concentration of active radicals decomposed from tetrahydrofuran. Figure8 and the black rectangles in Figure7 illustrate how sensitive is the ultrasonic effect to the ratio of THF/methanol and shows that the chemical treatment of the Si surface has an opposite effect in THF and THF/methanol. The decay time τ0 in a mixture of tetrahydrofuran and methanol given in Table1 behaves quite similarly under the chemical and sonochemical treatments. Therefore, slight variation of the THF concentration does not change appreciably the relative weights of the decay components, slightly varying the values of τ0 that might be related to a solvent polarity [57]. Increasing the volume of methanol in the mixture to the ratio of 1/1 gives the changes in the SPV amplitudes (see Figure8), which are rather similar to that observed in the 3/1 mixture (Figure7a) in cases of the chemical treatment (curve 2 in Figure8) and sonication at 28 kHz (curve 3). The higher-frequency treatment quenches the SPV signal, as shown by curve 4 in Figure8. In contrast, the decay time τ0 exhibits a decrease when treating chemically (from 33.7 to 13.4 µs) or sonochemically at 28 kHz (from 26.2 to 17.8 µs), while τ0 is enhanced by increasing the sonication frequency (from 19.2 to 24.9 µs in our measurements). Most significantly, the SPV amplitude is similarly enhanced in the 3/1 mixture upon sonication at both 28 and 400 kHz, as shown by black rectangles in Figure7a. Furthermore, the SPV amplitude is effectively quenched both chemically and sonochemically in pure methanol, as shown in Figure9a,b. Molecules 2021, 26, x FOR PEER REVIEW 12 of 16
Molecules 2021,, 26,, x 3756 FOR PEER REVIEW 1212 of of 16
Figure 8. Time-dependent SPV of Si samples before (curve 1) and after chemical (curve 2) and sonochemical treatments at about 28 (curve 3) and 400 kHz (curve 4) in tetrahydrofuran/methanol mixture with a volume ratio of 1/1.
Most significantly, the SPV amplitude is similarly enhanced in the 3/1 mixture upon sonication at both 28 and 400 kHz, as shown by black rectangles in Figure 7a. Further-
Figuremore, 8.theTime-dependent SPV amplitude SPV is effectively of Si samples quenched before (curve both chemically 1) and after and chemical sonochemically (curve 2) and in Figure 8. Time-dependent SPV of Si samples before (curve 1) and after chemical (curve 2) and sonochemicalpure methanol treatments, as shown at about in Figure 28 (curve 9a,b. 3) and 400 kHz (curve 4) in tetrahydrofuran/methanol sonochemical treatments at about 28 (curve 3) and 400 kHz (curve 4) in tetrahydrofuran/methanol mixture with a volume ratio of 1/1. mixture with a volume ratio of 1/1.
Most significantly, the SPV amplitude is similarly enhanced in the 3/1 mixture upon sonication at both 28 and 400 kHz, as shown by black rectangles in Figure 7a. Further- more, the SPV amplitude is effectively quenched both chemically and sonochemically in pure methanol, as shown in Figure 9a,b.
(a) (b)
FigureFigure 9. 9.( a()a Time-dependent) Time-dependent SPV SPV of of Si samplesSi samples before before (curve (curve 1) and 1) and after after chemical chemical (curve (curve 2) and 2) sonochemical and sonochemical treatments treat- atments about at 28 about (curve 28 3)(curve and 400 3) and kHz 400 (curve kHz 4) (curve in methanol; 4) in methanol (b) relative; (b) r changeelative inchange the SPV in the amplitudes SPV amplitudes due to exposure due to exposure of Si to methanolof Si to methanol and sonication and sonication in methanol in methanol at about at 28 about and 40028 and kHz. 400 The kHz. data The in data (b) are in ( obtainedb) are obtained from decays from decays shown shown in (a) at in (a) at time moment 푡 = 0.
time moment t = 0.
The essence of these observations lies in the fact that methanol itself can probably be (a) The essence of these observations lies in the fact that methanol(b) itself can probably be decomposeddecomposed[ [5858]] ultrasonically: ultrasonically: Figure 9. (a) Time-dependent SPV of Si samples before (curve 1) and after chemical (curve 2) and sonochemical treat- ments at about 28 (curve 3) and 400 kHz (curve 4) in methanol; (bCH) relative3COHH3OH change→→2H22H in+2 the+COCO SPV amplitudes due to exposure (26)(26) of Si to methanol and sonication in methanol at about 28 and 400 kHz. The data in (b) are obtained from decays shown in (a) at time moment 푡 = 0. The difference between sonication in tetrahydrofuran and methanol might therefore be caused by the occurrence of the atomic hydrogen released from THF in the reaction givenThe by essence Equation of (22),these which observations provides lies an effectivein the fact passivation that methanol of the itself Si surface. can pro Basedbably onbe decomposedthe fact that the[58] H-bonded ultrasonically methanol: dimer is formed [59,60] and succeeded in decreasing passivation ability of hydrogen, a slight decrease in the SPV signal in THF/methanol CH OH → 2H + CO (26) observed in Figure7a at Sono 400 kHz can3 naturally2 be explained.
Molecules 2021, 26, 3756 13 of 16
The interesting feature is the ultrasonic frequency effect observed in Figure7a for the red and black rectangles. Such an effect might be driven by the complex dissociation dynamics in THF and THF/methanol mixtures [61–63]. It may be suggested that some kind decay process with hydrogen happens in the time window between 36 µs (corresponding to 28 kHz) and 2.5 µs (400 kHz). Another major effect of sonication in methanol, which is most clearly seen in curve 4 of Figure9a (also enlarged in the inset) taken at about 400 kHz, is a complicated shape of the decay. This seemingly stems from the fact that this SPV decay can be decomposed into two transients with positive and negative values of the partial amplitude factors U01 and U02. This suggests that, upon the higher-frequency sonication in methanol, there is a competitive path of separating photoexcited electrons and holes on the surface, which is heavily blocked by lowering the sonication frequency. The origin of this behavior is not precisely determined and a more detailed examination of it, correlating the decay shape transformation with the sonication frequency and testing whether these data are mutually consistent, is warranted.
4. Conclusions Here, we sonicated tetrahydrofuran, which can act as a source of hydrogen capable of improving the photovoltaic response of Si wafers. Two ultrasonic frequencies of about 28 and 400 kHz were employed. The enhanced SPV signal was observed due to increasing sonication frequency. Thus, the SPV response was increased up to ≈80% due to sonication at 400 kHz, whereas the signal decreased down to ≈75% of the initial value when the frequency was lowered to about 28 kHz. The addition of methanol was found to suppress the significant difference in the sonication effect at the two frequencies. Notwithstanding the many scenarios that might be involved in properly interpreting the observed frequency behavior, this work hinted at the possibility of involvement of atomic hydrogen decomposed from tetrahydrofuran. This is known as effective passivator of surface silicon dangling bonds. It is therefore suggested that different decay processes with hydrogen occur over a physically meaningful time scale from 2.5 to 36 µs, which corresponds to the ultrasonic frequencies of 400 and 28 kHz, respectively.
Author Contributions: Conceptualization, O.K. and V.S.; methodology, A.P. and A.N.; software, A.N.; validation, A.P., A.N., and O.K.; formal analysis, A.P., A.N., O.K., and V.S.; investigation, A.P. and A.N.; resources, O.K. and V.S.; data curation, A.P. and O.K.; writing—original draft preparation, O.K.; writing—review and editing, O.K. and V.S.; visualization, A.P., A.N., and O.K.; supervision, V.S.; project administration, O.K. and V.S.; funding acquisition, V.S. All authors have read and agreed to the published version of the manuscript. Funding: The work at Kyiv was funded by the Ministry of Education and Science of Ukraine, grant number 0119U100303. Financial support from the University of Vienna is also acknowledged. Open Access Funding by the University of Vienna. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Conflicts of Interest: The authors declare no conflict of interest. Sample Availability: Samples of the compounds are not available from the authors.
References 1. Vinatoru, M.; Mason, T. Jean-Louis Luche and the Interpretation of Sonochemical Reaction Mechanisms. Molecules 2021, 26, 755. [CrossRef] 2. Suslick, K.S. Sonochemistry. Science 1990, 247, 1439–1445. [CrossRef] 3.S áez, V.; Mason, T.J. Sonoelectrochemical Synthesis of Nanoparticles. Molecules 2009, 14, 4284–4299. [CrossRef] Molecules 2021, 26, 3756 14 of 16
4. Xu, H.; Zeiger, B.W.; Suslick, K.S. Sonochemical synthesis of nanomaterials. Chem. Soc. Rev. 2013, 42, 2555–2567. [CrossRef] [PubMed] 5. Gedanken, A.; Perelshtein, I. Power ultrasound for the production of nanomaterials. In Power Ultrasonics. Applications of High-Intensity Ultrasound; Gallego-Juárez, J.A., Graff, K.F., Eds.; Woodhead Publishing: Amsterdam, The Netherlands, 2015; pp. 543–576. [CrossRef] 6. Taha, A.; Ahmed, E.; Ismaiel, A.; Ashokkumar, M.; Xu, X.; Pan, S.; Hu, H. Ultrasonic emulsification: An overview on the preparation of different emulsifiers-stabilized emulsions. Trends Food Sci. Technol. 2020, 105, 363–377. [CrossRef] 7. Sandhya, M.; Ramasamy, D.; Sudhakar, K.; Kadirgama, K.; Harun, W. Ultrasonication an intensifying tool for preparation of stable nanofluids and study the time influence on distinct properties of graphene nanofluids—A systematic overview. Ultrason. Sonochem. 2021, 73, 105479. [CrossRef] 8. Maisonhaute, E.; Prado, C.; White, P.C.; Compton, R.G. Surface acoustic cavitation understood via nanosecond electrochemistry. Part III: Shear stress in ultrasonic cleaning. Ultrason. Sonochem. 2002, 9, 297–303. [CrossRef] 9. Cobley, A.; Mason, T.; Cobley, A.; Mason, T.J. The evaluation of sonochemical techniques for sustainable surface modification in electronic manufacturing. Circuit World 2007, 33, 29–34. [CrossRef] 10. Paniwnyk, L.; Cobley, A. Ultrasonic Surface Modification of Electronics Materials. Phys. Procedia 2010, 3, 1103–1108. [CrossRef] 11. Arruda, L.B.; Orlandi, M.; Lisboa-Filho, P. Morphological modifications and surface amorphization in ZnO sonochemically treated nanoparticles. Ultrason. Sonochem. 2013, 20, 799–804. [CrossRef] 12. Savkina, R.K.; Gudymenko, A.I.; Kladko, V.P.; Korchovyi, A.; Nikolenko, A.S.; Smirnov, A.; Stara, T.; Strelchuk, V.V. Silicon Substrate Strained and Structured via Cavitation Effect for Photovoltaic and Biomedical Application. Nanoscale Res. Lett. 2016, 11, 183. [CrossRef][PubMed] 13. Skorb, E.V.; Möhwald, H. Ultrasonic approach for surface nanostructuring. Ultrason. Sonochem. 2016, 29, 589–603. [CrossRef] 14. Bai, F.; Wang, L.; Saalbach, K.-A.; Twiefel, J. A Novel Ultrasonic Cavitation Peening Approach Assisted by Water Jet. Appl. Sci. 2018, 8, 2218. [CrossRef] 15. Nadtochiy, A.; Korotchenkov, O.; Schlosser, V. Sonochemical Modification of SiGe Layers for Photovoltaic Applications. Phys. Status Solidi A 2019, 216, 1900154. [CrossRef] 16. Angermann, H.; Henrion, W.; Roseler, A. Wet-chemical conditioning of silicon: Electronic properties correlated with the surface morphology. In Silicon-Based Materials and Devices; Nalwa, H.S., Ed.; Academic Press: San Diego, CA, USA, 2001; Volume 1, pp. 267–298. 17. Fuchs, F. Ultrasonic cleaning and washing of surfaces. In Power Ultrasonics. Applications of High-Intensity Ultrasound; Gallego- Juárez, J.A., Graff, K.F., Eds.; Woodhead Publishing: Amsterdam, The Netherlands, 2015; pp. 577–609. [CrossRef] 18. Yasui, K. Influence of ultrasonic frequency on multibubble sonoluminescence. J. Acoust. Soc. Am. 2002, 112, 1405–1413. [CrossRef] [PubMed] 19. Wu, X.; Joyce, E.M.; Mason, T.J. Evaluation of the mechanisms of the effect of ultrasound on Microcystis aeruginosa at different ultrasonic frequencies. Water Res. 2012, 46, 2851–2858. [CrossRef][PubMed] 20. Joyce, E.; Phull, S.; Lorimer, J.; Mason, T.J. The development and evaluation of ultrasound for the treatment of bacterial suspensions. A study of frequency, power and sonication time on cultured Bacillus species. Ultrason. Sonochem. 2003, 10, 315–318. [CrossRef] 21. Koda, S.; Miyamoto, M.; Toma, M.; Matsuoka, T.; Maebayashi, M. Inactivation of Escherichia coli and Streptococcus mutans by ultrasound at 500kHz. Ultrason. Sonochem. 2009, 16, 655–659. [CrossRef] 22. Calimli, M.H.; Nas, M.S.; Acidereli, H.; Sen, F. Sonochemical methods and their leading properties for chemical synthesis. In Green Sustainable Process for Chemical and Envi-Ronmental Engineering and Science, 1st ed.; Asiri, A.M., Kanchi, S., Eds.; Elsevier: Amsterdam, The Netherlands, 2019; Chapter 13; pp. 355–365. 23. Bussemaker, M.J.; Zhang, D. A phenomenological investigation into the opposing effects of fluid flow on sonochemical activity at different frequency and power settings. 1. Overhead stirring. Ultrason. Sonochem. 2014, 21, 436–445. [CrossRef] 24. Monnier, H.; Wilhelm, A.; Delmas, H. The influence of ultrasound on micromixing in a semi-batch reactor. Chem. Eng. Sci. 1999, 54, 2953–2961. [CrossRef] 25. Van Der Sluis, L.W.M.; Versluis, M.; Wu, M.K.; Wesselink, P.R. Passive ultrasonic irrigation of the root canal: A review of the literature. Int. Endod. J. 2007, 40, 415–426. [CrossRef] 26. Dashtimoghadam, E.; Johnson, A.; Fahimipour, F.; Malakoutian, M.; Vargas, J.; Gonzalez, J.; Ibrahim, M.; Baeten, J.; Tayebi, L. Vibrational and sonochemical characterization of ultrasonic endodontic activating devices for translation to clinical efficacy. Mater. Sci. Eng. C 2020, 109, 110646. [CrossRef][PubMed] 27. Podolian, A.; Nadtochiy, A.; Kuryliuk, V.; Korotchenkov, O.; Schmid, J.; Drapalik, M.; Schlosser, V. The potential of sonicated water in the cleaning processes of silicon wafers. Sol. Energy Mater. Sol. Cells 2011, 95, 765–772. [CrossRef] 28. Mason, T.J.; Lorimer, J.P. Applied Sonochemistry: Uses of Power Ultrasound in Chemistry and Processing; Wiley: Weinheim, Germany, 2002. 29. Son, Y.; No, Y.; Kim, J. Geometric and operational optimization of 20-kHz probe-type sonoreactor for enhancing sonochemical activity. Ultrason. Sonochem. 2020, 65, 105065. [CrossRef] 30. Ensminger, D.; Bond, L.J. Ultrasonics Fundamentals, Technologies, and Applications, 3rd ed.; CRC Press: Boca Raton, FL, USA, 2012. 31. Ostrovskii, I.V.; Nadtochiy, A.B. Domain resonance in two-dimensional periodically poled ferroelectric resonator. Appl. Phys. Lett. 2005, 86, 222902. [CrossRef] Molecules 2021, 26, 3756 15 of 16
32. Royer, D.; Dieulesaint, E. Elastic Waves in Solids I, 1st ed.; Springer: Berlin, Germany, 2010. 33. Nadtochiy, A.; Shmid, V.; Korotchenkov, O. Miniature ultrasonic transducer for lab-on-a-chip applications. In Proceedings of the 2020 IEEE 40th International Conference on Electronics and Nanotechnology (ELNANO), Kyiv, Ukraine, 22–24 April 2020; Institute of Electrical and Electronics Engineers (IEEE): New York, NY, USA, 2020; pp. 425–429. 34. Koda, S.; Kimura, T.; Kondo, T.; Mitome, H. A standard method to calibrate sonochemical efficiency of an individual reaction system. Ultrason. Sonochem. 2003, 10, 149–156. [CrossRef] 35. Barchouchi, A.; Molina-Boisseau, S.; Gondrexon, N.; Baup, S. Sonochemical activity in ultrasonic reactors under heterogeneous conditions. Ultrason. Sonochem. 2021, 72, 105407. [CrossRef][PubMed] 36. Yang, X.-Y.; Wei, H.; Li, K.-B.; He, Q.; Xie, J.-C.; Zhang, J.-T. Iodine-enhanced ultrasound degradation of sulfamethazine in water. Ultrason. Sonochem. 2018, 42, 759–767. [CrossRef] 37. Jovanovski, V.; Orel, B.; Ješe, R.; Vuk, A.Š.; Mali, G.; Hocˇcevar, S.B.; Grdadolnik, J.; Stathatos, E.; Lianos, P. Novel Polysilsesquioxane−I-/I3-Ionic Electrolyte for Dye-Sensitized Photoelectrochemical Cells. J. Phys. Chem. B 2005, 109, 14387–14395. [CrossRef][PubMed] 38. Apostolopoulou, A.; Margalias, A.; Stathatos, E. Functional quasi-solid-state electrolytes for dye sensitized solar cells prepared by amine alkylation reactions. RSC Adv. 2015, 5, 58307–58315. [CrossRef] 39. Nomura, S.; Mukasa, S.; Kuroiwa, M.; Okada, Y.; Murakami, K. Cavitation Bubble Streaming in Ultrasonic-Standing-Wave Field. Jpn. J. Appl. Phys. 2005, 44, 3161–3164. [CrossRef] 40. Yasuda, K.; Nguyen, T.T.; Asakura, Y. Measurement of distribution of broadband noise and sound pressures in sonochemical reactor. Ultrason. Sonochem. 2018, 43, 23–28. [CrossRef] 41. Jamshidi, R.; Pohl, B.; Peuker, U.; Brenner, G. Numerical investigation of sonochemical reactors considering the effect of inhomogeneous bubble clouds on ultrasonic wave propagation. Chem. Eng. J. 2012, 189-190, 364–375. [CrossRef] 42. Sasmal, S.; Goud, V.V.; Mohanty, K. Ultrasound Assisted Lime Pretreatment of Lignocellulosic Biomass toward Bioethanol Production. Energy Fuels 2012, 26, 3777–3784. [CrossRef] 43. Ashokkumar, M. The characterization of acoustic cavitation bubbles—An overview. Ultrason. Sonochem. 2011, 18, 864–872. [CrossRef][PubMed] 44. Hunger, R.; Fritsche, R.; Jaeckel, B.; Jaegermann, W.; Webb, L.J.; Lewis, N.S. Chemical and electronic characterization of methyl-terminated Si(111) surfaces by high-resolution synchrotron photoelectron spectroscopy. Phys. Rev. B 2005, 72, 045317. [CrossRef] 45. Johansson, E.; Hurley, P.T.; Brunschwig, B.S.; Lewis, N.S. Infrared Vibrational Spectroscopy of Isotopically Labeled Ethyl- Terminated Si(111) Surfaces Prepared Using a Two-Step Chlorination/Alkylation Procedure. J. Phys. Chem. C 2009, 113, 15239–15245. [CrossRef] 46. Podolian, A.; Kozachenko, V.; Nadtochiy, A.; Borovoy, N.; Korotchenkov, O. Photovoltage transients at fullerene-metal interfaces. J. Appl. Phys. 2010, 107, 93706. [CrossRef] 47. Angermann, H. Characterization of wet-chemically treated silicon interfaces by surface photovoltage measurements. Anal. Bioanal. Chem. 2002, 374, 676–680. [CrossRef][PubMed] 48. Kerr, M.J.; Cuevas, A. Very low bulk and surface recombination in oxidized silicon wafers. Semicond. Sci. Technol. 2001, 17, 35–38. [CrossRef] 49. Gaubas, E.; Simoen, E.; Vanhellemont, J. Review—Carrier Lifetime Spectroscopy for Defect Characterization in Semiconductor Materials and Devices. ECS J. Solid State Sci. Technol. 2016, 5, P3108–P3137. [CrossRef] 50. Podolian, A.; Nadtochiy, A.; Korotchenkov, O.; Romanyuk, B.; Melnik, V.; Popov, V. Enhanced photoresponse of Ge/Si nanostruc- tures by combining amorphous silicon deposition and annealing. J. Appl. Phys. 2018, 124, 095703. [CrossRef] 51. Shmid, V.; Podolian, A.; Nadtochiy, A.; Yazykov, D.; Semenko, M.; Korotchenkov, O. Photovoltaic Characterization of Si and SiGe Surfaces Sonochemically Treated in Dichloromethane. J. Nano Electron. Phys. 2020, 12, 1023. [CrossRef] 52. Royea, W.J.; Juang, A.; Lewis, N.S. Preparation of air-stable, low recombination velocity Si(111) surfaces through alkyl termination. Appl. Phys. Lett. 2000, 77, 1988–1990. [CrossRef] 53. Klute, C.H.; Walters, W.D. The Thermal Decomposition of Tetrahydrofuran. J. Am. Chem. Soc. 1946, 68, 506–511. [CrossRef] 54. Lifshitz, A.; Bidani, M.; Bidani, S. Thermal reactions of cyclic ethers at high temperatures. Part 3. Pyrolysis of tetrahydrofuran behind reflected shocks. J. Phys. Chem. 1986, 90, 3422–3429. [CrossRef] 55. Verdicchio, M.; Sirjean, B.; Tran, L.-S.; Glaude, P.-A.; Battin-Leclerc, F. Unimolecular decomposition of tetrahydrofuran: Carbene vs. diradical pathways. Proc. Combust. Inst. 2015, 35, 533–541. [CrossRef] 56. Schmidt, J.; Peibst, R.; Brendel, R. Surface passivation of crystalline silicon solar cells: Present and future. Sol. Energy Mater. Sol. Cells 2018, 187, 39–54. [CrossRef] 57. Gude, C.; Rettig, W. Radiative and Nonradiative Excited State Relaxation Channels in Squaric Acid Derivatives Bearing Differently Sized Donor Substituents: A Comparison of Experiment and Theory. J. Phys. Chem. A 2000, 104, 8050–8057. [CrossRef] 58. Hokenek, S.; Kuhn, J.N. Methanol Decomposition over Palladium Particles Supported on Silica: Role of Particle Size and Co-Feeding Carbon Dioxide on the Catalytic Properties. ACS Catal. 2012, 2, 1013–1019. [CrossRef] 59. Huisken, F.; Kulcke, A.; Laush, C.; Lisy, J.M. Dissociation of small methanol clusters after excitation of the O–H stretch vibration at 2.7 µ. J. Chem. Phys. 1991, 95, 3924–3929. [CrossRef] Molecules 2021, 26, 3756 16 of 16
60. Chowdhury, P. Infrared depletion spectroscopy suggests fast vibrational relaxation in the hydrogen-bonded aniline– tetrahydrofuran (C6H5–NH2 . . . OC4H8) complex. Chem. Phys. Lett. 2000, 319, 501–506. [CrossRef] 61. Janeckova, R.; May, O.; Milosavljevi´c,A.; Fedor, J. Partial cross sections for dissociative electron attachment to tetrahydrofuran reveal a dynamics-driven rich fragmentation pattern. Int. J. Mass Spectrom. 2014, 365-366, 163–168. [CrossRef] 62. De Bruycker, R.; Tran, L.-S.; Carstensen, H.-H.; Glaude, P.-A.; Monge, F.; Alzueta, M.U.; Battin-Leclerc, F.; Van Geem, K.M. Experi- mental and modeling study of the pyrolysis and combustion of 2-methyl-tetrahydrofuran. Combust. Flame 2017, 176, 409–428. [CrossRef] 63. Kaur, S.P.; Ramachandran, C. Hydrogen-tetrahydrofuran mixed hydrates: A computational study. Int. J. Hydrogen Energy 2018, 43, 19559–19566. [CrossRef]