Laser-controlled field effect in graphene/hexagonal boron nitride heterostructures
Igor Wlasny, Roman Stepniewski, Zbigniew Klusek, Wlodzimierz Strupinski, Andrzej Wysmolek
aa r X i v : . [ c ond - m a t . m t r l - s c i ] M a y Laser-controlled field effect in graphene/hexagonal boron nitrideheterostructures
I. Wlasny, a) R. Stepniewski, Z. Klusek, W. Strupinski,
4, 5 and A. Wysmolek Institute of Experimental Physics, Faculty of Physics, University of Warsaw, Pasteura 5, 02-093 Warsaw,Poland Institute of Experimental Physics, Faculty of Physics, University of Warsaw, Pasteura 5,02-093 Warsaw Poland Department of Solid State Physics, Faculty of Physics and Applied Informatics, University of Lodz,Pomorska 149/153, 90-236 Lodz, Poland Institute of Electronic Materials Technology, Wolczynska 133, 01-919 Warsaw,Poland Faculty of Physics, Warsaw University of Technology (WUT), Koszykowa 75, 00-662 Warsaw,Poland (Dated: 17 May 2018; Revised 17 May 2018)
The possibility of modification of the local properties of hexagonal boron nitride (h-BN) by laser irradiationis investigated. Investigations conducted using both Raman spectroscopy and electrostatic force microscopywere performed. Laser light induced modifications are found to cause no structural changes. However, theyimpact the Raman spectra and local charge state of the material. They are also shown to be stable in timeand during electrical grounding of the sample. A mechanism of photoionization of deep defects present inh-BN is proposed to explain the observed phenomenon. The discussed effect opens up a new method ofnanostructuration of h-BN based planar heterostructures.
CONTENTS
I. Introduction II. Experimental details
III. Results and discussion
IV. Conclusions V. Supplementary Material VI. Acknowledgments VII. References I. INTRODUCTION
Two-dimensional van der Waals heterostructures, inrecent years, attract growing attention of the scientificand industry communities . They owe their popularityto the possibility of creating planar nanodevices, which a) Electronic mail: [email protected] is related to the characteristics of van der Waals mate-rials as well as the possibility of tailoring their electricaland optical properties . This shows their high poten-tial for application in various electronic or optoelectronicdevices. In particular, one of the most thoroughly investi-gated heterostructures, composed of graphene and hexag-onal boron nitride, finds use in e.g. tunneling diodes ortransistors .Graphene is a carbon allotrope with atoms arranged ina hexagonal lattice . From the point of view of its elec-tronic structure it is a zero-bandgap semiconductor withexceptionally high charge carrier mobility . However,in relation to the heterostructure construction, the mostimportant aspect of graphene is that its properties aresensitive to a broad spectrum of factors, such as growthmethod , interaction with a substrate or the environ-mental interactions, such as an electric field effect .Hexagonal boron nitride (h-BN) has a similar struc-ture to that of graphene , including the lattice configu-ration as well as a closely matched lattice constant. It is,however, a semiconductor with a wide bandgap of over5 eV . As an ultra-flat dielectric, free of surface chargenon-uniformities , it interacts weakly with graphenelayers within a heterostructure. Due to aforementionedslight lattice mismatch and possible rotation periodic su-perstructures with new functionalities could be formed .In this article we investigate the possibility of control-ling the 2D materials by means of modification of the h-BN charge state using a focused light beam. In our studywe have investigated h-BN samples both exfoliated frombulk crystals as well as transferred CVD-grown (CVD -chemical vapor deposition). Our Raman scattering (RS)and electrostatic force microscopy (EFM) investigationsindicate the possibility of locally influencing the prop-erties of the hexagonal boron nitride as well as provethat laser-based experimental techniques should be usedwith caution when investigating two-dimensional mate-rials. Our studies also allowed us to pinpoint the originof the observed changes as well as to study their dynam-ics. Our investigations of graphene/h-BN heterostruc-ture show that the photoionization of deep defect cen-ters has an effect on the charge carrier concentration ofgraphene due to a field-effect. This is particularly im-portant, as the same effect occurs within heterostructuresbased on other 2D materials. The discussed phenomenonopens up a new method of nanostructuration of planarheterostructures based on h-BN as well as provides newpossibilities of graphene/h-BN structure applications. II. EXPERIMENTAL DETAILSA. Sample preparation
The investigations presented in this article were con-ducted on a series of samples, allowing us to investigateboth the effect of laser illumination on the characteris-tics of standalone h-BN layers and graphene/h-BN het-erostructures, which were used to investigate the effectof h-BN modification on graphene.The first sample, referred to as h-BN/SiO through-out the text, was prepared by mechanical exfoliation ofcommercially available h-BN single crystals . The ini-tial exfoliation was performed using a standard dicingtape (Microworld M07). The final exfoliation and trans-fer onto a target substrate was performed with 0.1 mmthick PDMS (polydimethylosiloxane) film. Flakes weredeposited onto a boron doped (p-type) silicon substrate(resistivity of 1Ω cm) with 90 nm layer of the thermal ox-ide. Such substrates enhance the optical contrast of theh-BN flakes allowing for their quick identification . ThePDMS film was removed from SiO during the coolingphase after previously heating the sample up to 80 ◦ C,which increased the effectiveness of the deposition .The sample referred to as h-BN/G/SiC served as amodel for a heterostructure of graphene and hexago-nal boron nitride. It was prepared using the[previouslydescribed exfoliation method. However, instead ofa SiO /Si substrate, a 4H-SiC crystal with epitaxialgraphene monolayer has been used. The graphene layerwas grown using a commercial horizontal CVD hot-wallreactor with propane gas as a precursor .The last sample studied in this article (G/h-BN/SiO )was prepared by the deposition of graphene and hexag-onal boron nitride onto the SiO /Si substrate. Bothgraphene and h-BN were synthesized using the CVDmethod on high purity Cu foil substrates. They weretransferred onto SiO with a modified wet-transfermethod . Before the etching of copper, a 0.1 mm thickPDMS layer was deposited on the h-BN/Cu. The etch-ing was performed in 0.1 m aqueous solution of iron(III)nitrate nonahydrate at 21 ◦ C over a time span of 20hours. PDMS with h-BN/graphene layers was cleaned of the iron(III) nitrate contamination with ten cycles ofbathing in deionized water. Subsequently, PDMS withh-BN/graphene was carefully placed on the SiO sub-strate, where it was pressed down for 30 minutes. ThePDMS layer was removed during the cooling phase afterpreviously heating the sample to 80 ◦ C. B. Atomic/Electrostatic Force Microscopy
Both atomic force microscopy (AFM) and electrostaticforce microscopy (EFM) measurements were performedusing NT-MDT Ntegra Aura microscope working in at-mospheric conditions (air temperature 21 ◦ C, pressure1000 hPa), with NT-MDT NSG10/Au cantilevers coatedwith 35 nm layer of Au. The setup was connected elec-trically with the sample to provide a positive bias on thecantilever. Such setup allowed both for imaging the to-pography of the samples in the non-contact mode andEFM measurements. Each of the images was collectedwith 256 x 256 pts resolution. The results were analyzedusing the Gwyddion 2.40 software . C. Raman Spectroscopy and Optical Microscopy
Sample illumination, Raman measurements and Opti-cal Microscopy were performed using a Renishaw inViasystem equipped with Olympus MPLN 100x objectivewith 100x magnification and and automated XYZ trans-lation stage with 100 nm spatial resolution. A RenishawRL532C50 single mode laser with a nominal 45 mW out-put power has been used as an excitation source. Coupledwith the optical microscope it allows to create a focusedspot with the measured power of 13 mW and diame-ter of about 500 nm on the surface of the sample. Thissetup allowed the acquisition of optical images as well ashigh-resolution Raman measurements and provided theillumination source to the selected areas of the sample.The obtained Raman spectra were analyzed by numer-ical fitting of the model spectrum based on Lorentziancurves using the Wolfram Mathematica 11.2 software.
III. RESULTS AND DISCUSSIONA. Electric field measurements
The results presented in this section were obtained onthe h-BN/SiO sample. This part of the experiment wasperformed in order to confirm that the modification isrelated to the emergence of local electric fields in thesample. To this end a h-BN fragment with large areaflat surface was selected. Its optical microscopy image isshown in Figure 1 a). Basing on the optical contrast, thethickness of the sample is about 5 nm on the terraces withseveral thicker topographical features . The thickness a) optical imagec) AFM topography b) laser illumination sited) EFM +10V532 nm13 mW3600 s FIG. 1. a) Optical image of the h-BN flake on h-BN/SiO sample, b) Schematic view of the flake with path of the laserillumination indicated with a green line, c) AFM topographyof the h-BN flake, d) EFM image of the sample after illumina-tion acquired with 10 V positive bias between AFM cantileverand sample. and surface characteristics are further confirmed by AFMtopography measurements shown in Figure 1 c).The sample was illuminated in a linear series of points,as indicated in Figure 1 b), with conditions that, as webelieve, would maximize the observed modification of thelocal characteristics of the h-BN fragment. Each of thepoints was separated by 500 nm distance. They wereilluminated with 532 nm laser with 13 mW power and aspot of about 500 nm diameter (power density of about16.5 mW µ m -2 ) over the time span of 3600 seconds.The topography presented in Figure 1 c) was acquiredafter the illumination process. It indicates some topo-graphical features that were not seen during the opticalmeasurements - few wrinkles of the h-BN material andslight amount of residues left after the deposition pro-cess. However, no distinct features can be found wherethe illumination was conducted. This clearly shows thatdespite using the high-power density radiation the sam-ple has not been damaged and shows no visible signsof degradation or deposition of the contamination on thesurface. The correlation between the sample illuminationand the electric field emergence, however, can be clearlyseen in EFM measurements, as presented in Figure 1 d).In the image a bright area can be seen where the samplewas illuminated indicating the presence of electric inter-action between the h-BN and the cantilever. This clearlyproves that incident focused laser light onto hexagonalboron nitride causes changes in the charge carrier dis-tribution within this material and leads to emergence oflocal electric fields. As shown later in this article, thiseffect may be used to control the characteristics of other2D materials present in the vicinity of the modified h-BN by means of the field-effect with non-invasive method ofillumination with high-power density light. Similar effectcan also be found after illumination with shorter timesand lower power densities (see supplementary material). B. Time-resolved Raman measurements
FIG. 2. Changes of a, c) position and b, d) FWHM (fullwidth at half maximum) of E g Raman line of h-BN duringthe illumination with 532 nm laser with of a, b) 1.3 mW powerand c, d) 0.13 mW power respectively and 0.5 µ m spot size. Purple, dashed line and orange, dashed line presented in a)and b) present the theoretical model fitted to the results. e)the schematics of the electron transitions between the defectcenters in illuminated h-BN assumed in the theoretical modeldescribing the observed photo-ionization of the h-BN. Results presented in the previous chapter proved thatthe laser illumination of the hexagonal boron nitrideleads to the emergence of local electric fields. However,in order to fully utilize this phenomenon to control othervan der Waals materials within the heterostructure itis important to identify the processes behind this phe-nomenon as well as to investigate the dynamics of the oc-curring changes. To this end in-situ Raman Spectroscopymeasurements were performed on the h-BN/SiO sample.The exfoliated flake was modified in several points witha focused beam of 532 nm laser. Neutral-density filterswere used in order to adjust the power density to the de-sired levels. The illuminating beam was used to inducethe Raman scattering effect for whole duration of themodification process. The time of acquisition for each ofthe spectra was adjusted to reach a satisfactory signal-to-noise ratio. Each of the spectra has been fitted with aLorentzian curve, which was used to describe the spectralline related to the E g vibrational mode of h-BN, whichis located at the Raman shift of about 1365 cm -126 . Themodel results are presented in Figure 2 a) and b).The changes of the position of the E g line are seenin Figure 2 a). The initial position of 1365.4 cm -1 isrelated to several factors, such as the thickness or its ini-tial charge state of the h-BN fragment . During the firsthour of the illumination the E g line is shifted towardshigher values, eventually reaching the peak value of about1365.55 cm -1 . After that point, the line is shifted down-wards to about 1365.3 cm -1 after 2 hours of the process.Subsequently, the line is shifted towards higher valuesagain, albeit at much slower rate.The changes can also be seen in the FWHM (full widthat half maximum) of the E g line (as shown in Figure 2b). This parameter starts at about 7.5 cm -1 and quicklyincreases to about 8.5 cm -1 after nearly 2 hours into theprocess. Next, the width of the spectral line seems to bereaching a stable level and does not undergo any changeswithin the time of observation.The described behavior can be explained by the mod-ifications of the charge state of deep defects which areinduced by the illumination with light. In the proposedmodel we assume the donors to be of a shallow typeand delocalized within the area of several lattice con-stants and, therefore, with limited coupling to the lattice.On the other hand, strongly localized deep acceptor cen-ters can substantially influence the h-BN structure uponthe charge transfer induced by photo-ionization. Thus,we associate the observed shifts of the Raman line en-ergy with the concentration of acceptors in the neutralstate .In a simplified model we assume the presence of defectlevels related to acceptors with concentration N A anddonors with concentration N D +N A . We assume that oursample is n-type and in thermodynamic equilibrium con-centration of ionized donors and acceptors are the sameand equal to N A . It was shown that Coulomb cou-pling and an additional short-range interaction promotethe tendency towards self-compensation by the formationof donor-acceptor pairs. Due to the Coulomb interac-tion the minimum energy of the system is reached whenthe donors nearest to the ionized acceptors are ionized.Therefore we divide the donor centers into two groups:”near” - located near acceptor states, initially ionized,with concentration N A and ”far” - initially neutral, lo-cated far from the acceptors, with concentration of N D .The incidence of light onto the crystal may change thisequilibrium state. In the simple model, with the excita-tion below the energy gap, we take into account the ex-citation that transforms the ionized donor-acceptor pairinto a neutral one with probability rate C (Figure 2 e)and excitations of neutral donors which create free elec- trons in the conduction band with the probability rateA, the same for both near and far donors (Figure 2 e)).As recombination processes we take into account donoracceptor recombination for near donors with the recombi-nation time τ da , as well as between far neutral donors andneutral acceptors with significantly bigger recombinationtime τ da . Free electrons from the conduction band can,within our model, either be trapped by ionized donorwith the recombination time τ cd or escape outside theilluminated part, with the escape time τ out . We apply alimit N Z to the concentration of electrons that can es-cape from the illuminated region. To describe this effectwe solve a set of rate equations for the described system dn nd ( t ) dt = Cn na ( t )( N A − n nd ( t )) − τ da n nd ( t )( N A − n na ( t )) − An nd ( t ) + 1 τ cd n ( t )( N A − n nd ( t ))(1) dn fd ( t ) dt = − τ da n fd ( t )( N A − n na ( t )) − An fd ( t )+1 τ cd n ( t )( N D − n fd ( t )) (2) dn na ( t ) dt = 1 τ da n fd ( t )( N A − n na ( t )) − Cn na ( t )( N A − n nd ( t )) + 1 τ da n nd ( t )( N A − n na ( t )) (3) dn ( t ) dt = An nd ( t ) − τ cd n ( t )( N A − n da ( t )) + An fd − τ cd n ( t )( N D − n fd ( t )) − τ out n ( t )( N Z − n out ( t )) (4) dn out ( t ) dt = 1 τ out n ( t )( N Z − n out ( t )) (5)where: n nd and n fd are the concentration of electronspresent on near and far donors respectively, n is the freeelectron concentration in the conduction band, n na is theconcentration of negatively ionized acceptors, that aretreated as the source of electrons that can be excited toa near donor. n out is the concentration of electrons thatescape from illuminated part of the sample.We assume that each excitation and recombinationprocess is proportional to the concentration of the occu-pied initial states and to the concentration of empty finalstates. We do not apply any limits to the concentrationof the empty states in the conduction band only. Consid-ering equations 1-5 it is possible to calculate the changesin the occupation in each of the considered states. Wesolved this set of equations numerically with the initialparameters presented in Tab. I. TABLE I. The initial parameters of the model.Parameter Initial valuen nd (0) 0n fd (0) N D n na (0) N A n(0) 0n out (0) 0 We assume that the position of the E g line (E(t)) istied strictly to the strain of the h-BN lattice. The re-distribution of the charge carriers leads to emergence ofthe non-uniformity of the electric fields within a crystal,which distort the lattice by means of piezoelectricity .We assume that the changes of the position of the E g line are related to the number of neutral acceptor cen-ters (N A -n na (t)), therefore the position of the line canbe expressed with equation 6 E ( t ) = a ( N A − n na ( t )) + E (6)where E is the position of the E g line in the initialstate and a is a scaling parameter. It is worth notingthat similar correlation between the illumination and thechanges in crystalline structure is found in other semi-conductors, such as GaN , GaAs or AlGaAs .The full width of half maximum of the E g line is re-lated to the local disorder in the lattice of the illuminatedh-BN. We assume the disorder is introduced primarily bythe presence of ionized defect centers ( N D − n nd ( t )) +( N D − n fd ( t )) + n na ( t ), which create non-uniform elec-tric fields in the crystal leading to rise of the disorder.This effect is lowered by the fact that the electron-dipoleinteraction depends on the distance between those twoobjects, thus the dipole contribution can be written as R n nd ( t )( N A − n na ( t )) N A , where R is the parameter correspond-ing to strength of the dipole contribution into the effect.Therefore the FWHM (Σ(t)) of the E g Raman line isgiven by equation 7Σ( t ) = c ( n nd ( t ) + n fd ( t ) + ( N A − n na ( t )) − R n nd ( t )( N A − n na ( t )) N A ) + Σ (7)where c is the scaling parameter and Σ is the initialFWHM of the line.The equations 6 and 7 were fitted simultaneously tothe data presented in figure 2 a) and b), with the re-sult presented in the same images. Based on this theestimated parameters are presented in Tab. IIOur simple model describes the experimental data rea-sonably well, with the exception to the beginning of theprocess. This may indicate that the initial state of thesystem was different than assumed. Furthermore, thenumber of possible levels associated with the defects inhexagonal boron nitride may be higher than what is taken TABLE II. The fitting parameters for the model fitted to thechanges of Raman line position and FWHM for illuminationwith 1.3 mW and 0.13 mW 532 nm laser.Parameter 1.3 mW, 0.5 µ m spot 0.13 mW, 0.5 µ m spotA [h -1 ] 0.23 0.016C [h -1 ] 0.86 -1 τ cd [h] 1.5 10.5 τ da [h] 0.62 8.7 τ da [h] 4.0 84 τ out [h] 2 28N A [arb.units] 1.5 1.5N D [arb.units] 8.5 8.5N Z [arb.units] 0.5 2 into account in the model . Both effects may explain thementioned discrepancy, particularly if the probabilities ofelectron transitions are high.Fitting to the data obtained during the illuminationwith 532 nm laser with power of 0.13 mW and spot sizeof 0.5 µ m over a time span of 24 hours, as presented infigure 2 c) and d) results in the values shown in Table II.Again, the divergence from the data can be seen forthe early stages of the observed process. It seems coun-terintuitive for the lifetimes, τ da and τ da in particular, tochange with the power density of the illuminating light.This fact can be explained, however, by the possibilitythat the electron transitions may be mediated by theconduction, valence bands or both. In this case the illu-mination may influence the concentration of the chargecarriers which are transported in the sample. Further-more, changes in illumination may result in a shift of thequasi Fermi level. While this model is simplified and doesnot take all of the factors into account, such as additionaldefect center levels, it does explain the major trends inthe observed parameters within the investigated scope ofthe power densities and, therefore, in our opinion, givesan insight on the basic processes behind the changes ob-served in the experimental results.While the threshold for the process is not seen in ourresults, there are more parameters that may have an in-fluence on the process. One of the key aspects is estab-lishing whether interaction between the h-BN layers isinfluencing the process. This is important both to theanalysis of the physical process itself and the applica-tions, where the thickness control issues is vital.In order to investigate that aspect we conducted insitu Raman measurements on the exfoliated h-BN flakewith areas of different thickness on a h-BN/SiO sample(see Figure 3 a) and a monolayer CVD h-BN fragment onG/h-BN/SiO sample. The changes of the position of theE g Raman line are presented in Figure 3 b). It is worthnoting that the initial state of the exfoliated flake inves-tigated here, in particular the thicker area, was differentfrom that presented in Figure 2, however the changesthat can be seen are similar to those previously reported
FIG. 3. a) Optical microscopy image of the investigated h-BN flake on h-BN/SiO sample with points points of Ramanmeasurements indicated with green and yellow dot, b) thedynamics of E g line position changes during the illuminationon h-BN of different thicknesses - 0.3 nm (CVD monolayerh-BN, sample G/h-BN/SiO ), 5 nm and 12 nm (sample h-BN/SiO ). - after 1 hour of illumination the peak position decreasesrapidly, and starts increasing slowly after reaching mini-mum value. The modification of the CVD (0.3 nm thick-ness) h-BN also seems to be progressing with similar dy-namics, however, the last phase is not seen, due to thelong acquisition times needed to reach that phase - thelocal maximum appears after almost 2 hours of the illu-mination indicating that the photo-excitation occurs ata slower rate in case of this sample. This effect is mostlikely related to the significant difference in thickness cou-pled with a low absorption coefficient or the interactionof the boron nitride layers with the substrate. C. Modification stability
The effect of photo-excitation of the charge carriers bythe electromagnetic radiation within the hexagonal boronnitride was previously investigated within the scope ofthe short-term effects, where the changes induced by lightwere shown not to be stable and disappear after stoppingthe illumination . The effect described in this article,however, shows signs of the stability - the electric contrastmeasurements (seen in Figure 1 d) were conducted fewhours after the illumination, proving that the changes inthe charge carrier distribution remain stable at least forthat duration. However, in order to establish whetherthe created modifications will still be seen on the sampleafter longer times we conducted measurements on theh-BN/SiO sample 5 months after the illumination.The EFM image of the sample after that time is shown FIG. 4. EFM images of the illuminated h-BN flake on h-BN/SiO sample after a) 5 months of time, b) electricalgrounding of the sample. in Figure 4 a). It is clearly seen that the contrast hasnot changed and slight differences between the imagesin Figure 4 a) and Figure 2 d) can be attributed to thedifferences in the exact state of the used AFM tip andenvironmental conditions. This shows that the energylevels attributed to the defect centers , the charge stateof which is influenced by the illuminations are deep andmay be used to effectively trap the charge for extremelylong times. This also proves that the field effect gen-erated by the investigated phenomenon may be used topermanently influence the heterostructure and may indi-cate the prospective applications of this effect where thestability is vital, such as the non-volatile memories or innanostructured devices.Additionally, attempts to restore the initial state of theilluminated h-BN were performed. This has been doneby performing the AFM scans of the surface with the tipwith Au coating in the contact mode of AFM. The tipwas grounded electrically. After such attempts, however,the EFM contrast can still be seen (see Figure 4 b). Thissuggests that the procedure did not have a significantimpact on the charge state of the material and observedeffects are related to the charge bound to deep defectcenters. D. Response of graphene to h-BN modification
The last part of the analysis of the phenomenon oflight-induced charge carrier photo-excitation in hexago-nal boron nitride is the investigation of the influence ofthe electric fields generated by the effect on graphenewithin the heterostructures composed of these two mate-rials. The analysis of that effect is basing on RamanSpectroscopy measurements. The spectra were gath-ered during the illumination conducted on both the het-erostructures and the graphene layers on h-BN/G/SiCand G/h-BN/SiO samples with 1.3 mW 532 nm laserwith 0.5 µ m spot size. Lorentzian curves were fitted tothe G and 2D bands of the Raman spectra of graphene.Analysis of the parameters of the curves provides in-formation on the charge carrier concentration and themechanical strain on the material in in case of lown-type doping (under 2 · cm -2 ) and biaxial strain,which we assume is the case, as no changes in corruga-tion were observed during AFM measurements (see Fig.1 c)) and we see no evidence of a non-uniform charac-ter of the observed changes. Within those bounds, theenergy of the G band depends on the electron concen-tration linearly, while 2D is independent from this value.The energies of G and 2D band are related to mechanicalstrain ( ǫ ) and electron concentration (n) by equations 8and 9 , where γ G and γ D are Gr¨uneisen parameters forG and 2D bands respectively. a =7.38 · cm is aparameter calculated by linear fitting of E G (n) functionfor electrostatic gating and E G and E D are the bandpositions of the undoped graphene for Raman measure-ments with 532 nm excitation wavelength. We assumedthe γ G /γ D to be equal to 0.71 . E D = E D − γ D · E D · ǫ (8) E G = E G − γ G · E G · ǫ + n · a (9) FIG. 5. Changes in the a, c) electron concentration andb, d) strain in graphene and graphene/h-BN heterostructurein a, b) h-BN/G/SiC and c, d) G/h-BN/SiO samples duringillumination with 532 nm 1.3 mW laser with 0.5 µ m. spotsize. The results of the measurements are shown in Figure5. The differences in the reaction of the graphene andthe heterostructures can be clearly seen at first glance.During the entirety of the illumination the strain of thegraphene outside the heterostructures, both in case of theepitaxial material on the native substrate (h-BN/G/SiC)and after the transfer (G/h-BN/SiO ) show nearly no re-action, as shown in Figure 5 b) and d). The only observedchanges are slight and occur within the first 0.2 hours incase of the transferred material. This is most likely as-sociated with the reaction of the non-relaxed interface to the changes in the charge state of hexagonal boronnitride. This also has an impact on the charge carrierconcentration in the graphene fragment on the surface.Increase of this parameter may allow the interface be-tween graphene and surface to relax, which is consistentwith the results seen in Figure 5 d). The change is veryslight, however, reaching only 0.01 %.The major differences appear, however, in the electrondoping of the investigated samples (see Figure 5 a) andc)). As was the case for the strain, the graphene outsideof the heterostructure shows nearly no reaction to thelaser illumination in both of the investigated samples.The exact value of the doping is related to the fact thatthe material was grown on different substrates leading tothe different initial value of charge carrier concentration.In case of the h-BN/G/SiC sample the concentration ingraphene was at the level of about 0.9 10 cm -2 (Fig-ure 5 a)), while within the heterostructure this value waschanging from 0 to 1.2 10 cm -2 with decreasing rate.It is worth pointing out that the initial doping of theheterostructure is significantly different from standalonegraphene (graphene outside the heterostructure) mostlikely due to the possibility of transfer of static chargesto the system during exfoliation and transfer of the h-BNlayers.The evolution of the charge carrier concentration inG/h-BN/SiO progresses similarly, however, there arefew key differences, as shown in Figure 5 c). First,the change in the electron doping of the heterostruc-ture is much smaller and is contained within the range of0.35 10 cm -2 to 0.70 10 cm -2 . This fact is understand-able considering that the h-BN layer is much thinner andtherefore the density of the defect centers is lower. Thegenerated field, is weaker and thus has lesser impact ongraphene. The second difference lies in the fact that theelectron concentration is decreasing after 0.6 hours. Asbefore that time the character of the changes is similar tothat in Figure 5 a). It is likely that the rate of changes isdifferent. Again, this can be attributed to the fact that inthe G/h-BN/SiO sample we are dealing with the mono-layer material, which is related to the lower maximumcharge carrier concentration in the defect levels.Basing on the above results, it can be clearly seen thatthe presence of the h-BN causes the modification of thecharge carrier concentration in graphene during the il-lumination with high power density light. As the reac-tion of the standalone graphene is significantly lower thechanges can be attributed to the emergence of the non-uniform electric fields in h-BN. The above results showthat the electron concentration can be changed in widerange (from 0 to 10 cm -2 ), therefore it shows that it ispossible to control the electric conductivity of the het-erostructure using the illumination with high power den-sity light. This opens up new possibilities of applicationof the effect in data storage or electronic component andcircuit engineering. Furthermore, as the photo-excitationeffect occurs in h-BN, it may also be used it in tailoringthe properties of other 2D materials. IV. CONCLUSIONS
In summary, it was shown that the illumination of thehexagonal boron nitride leads to the emergence of lo-cal electric fields. This effect is attributed to transitionsof the electric charge carriers between the levels asso-ciated with the structural defects. Those defect levelswere proven to be located deep within the band gap of h-BN. Because of this fact the light-induced modification ofthe charge carrier concentration was shown to be stablewithin long periods of time.The results presented in this article show that the elec-tric fields appearing in h-BN may be used to control theproperties of other 2D materials, which can be use to con-struct the heterostructures with hexagonal boron nitride.In particular, the presented results show the influence ofillumination on graphene/h-BN heterostructure, wherethe electric field may be used to control the electron con-centration in graphene.The described effect allows for easy and non-invasivetailoring of the electronic properties of two-dimensionalmaterials and may open new directions of research of thevan der Waals heterostructures and their applications.
V. SUPPLEMENTARY MATERIAL
See supplementary material for more records of illu-mination of both h-BN and graphene/h-BN heterostruc-tures as well as selected Raman spectra of the data pre-sented in this article.
VI. ACKNOWLEDGMENTS
This work was supported by National Science Cen-tre project granted on the basis of the decision num-ber DEC-2015/16/S/ST3/00451. The work was also fi-nancially supported by National Science Centre project2015/19/B/ST3/03142 as well as the European UnionSeventh Framework Program (Grant. No. 604391Graphene Flagship).
VII. REFERENCES H. Nematian, M. Moradinasab, M. Pourfath, M. Fathipour, andH. Kosina. J. Appl. Phys. 111, 093512 (2012). S. Lu and A. J. H. McGaughey. J. Appl. Phys. 121, 115103(2017). C.-H. Yeh, Z.-Y. Liang, Y.-C. Lin, T.-L. Wu, T. Fan, Y.-C. Chu,C.-H. Ma, Y.-C. Liu, Y.-H. Chu, K. Suenaga, and P.-W. Chi.ACS Appl. Mat. Interfaces 9, 36181 (2017). I. Wlasny, M. Rogala, P. Dabrowski, P. J. Kowalczyk, A. Busi-akiewicz, W. Kozlowski, L. Lipinska, J. Jagiello, M. Aksienionek,Z. Sieradzki, I. Krucinska, M. Puchalski, E. Skrzetuska, Z.Draczynski, and Z. Klusek. Mat. Chem. Phys. 181, 409 (2016). K. T. Winther and K. S. Thygesen. 2D Mat. 2D Materials 4,025059 (2017). M. Rogala, P. Dabrowski, P. Kowalczyk, I. Wlasny, W. Ko-zlowski, A. Busiakiewicz, I. Karaduman, L. Lipinska, J. Bara-nowski, and Z. Klusek. Carbon 103, 235-241 (2016). S. H. Lee, M. S. Choi, J. Lee, C. Ho Ra, X. Liu, E. Hwang, J. H.Choi, J. Zhong, W. Chen, and W. J. Yoo Appl. Phys. Lett. 104,053103 (2014). N. Petrone, T. Chari, I. Meric, L. Wang, K. L. Shepard, and J.Hone ACS Nano 9 8953 (2015). A. K. Geim and K. S. Novoselov. Nat. Mater. 6, 183 (2007). A. H. Castro Neto, N. M. R. Peres, K. S. Novoselov, and A. K.Geim. Rev. Mod. Phys. 81, 109 (2009). J. D. Buron, F. Pizzocchero, P. U. Jepsen, D. H. Petersen, J. M.Caridad, B. S. Jessen, T. J. Booth, and P. Bggild. Sci. Rep. 5,12305 (2015). Y. J. Shin, R. Stromberg, R. Nay, H. Huang, A. T. Wee, H. Yang,and C. S. Bhatia. Carbon 49, 4070 (2011). G. Giovannetti, P. A. Khomyakov, G. Brocks, V. M. Karpan, J.Van Den Brink, and P. J. Kelly. Phys. Rev. Lett. 101, 026803(2008). F Schwierz. Nat. Nanotech. 5, 487 (2010). M. L. Hu, J. L. Yin, C. X. Zhang, Z. Yu, and L. Z. Sun. J. Appl.Phys. 109, 073708 (2011). H. Jeong, S. Bang, H. M. Oh, H. J. Jeong, S. J. An, G. H. Han,H. Kim, K. K. Kim, J. C. Park, Y. H. Lee, G. Lerondel, and M.S. Jeong. ACS Nano 9, 10032 (2015). C. R. Dean, A. F. Young, I. Meric, C. Lee, L. Wang, S. Sorgenfrei,K. Watanabe, T. Taniguchi, P. Kim, K. L. Shepard, and J. Hone.Nat. Nanotech. 5, 722 (2010). K. K. Kim, A. Hsu, X. Jia, S. M. Kim, Y. Shi, M. Dresselhaus,T. Palacios, and J. Kong. Nano Lett. 12, 161166 (2012). M. Yankowitz, J. Xue, D. Cormode, J. D. Sanchez-Yamagishi,K. Watanabe, T. Taniguchi, P. Jarillo-Herrero, P. Jacquod, andB. J. LeRoy. Nat. Phys. 8, 382386 (2012). K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang,S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov. Science,306,666 (2004). R. V. Gorbachev, I. Riaz, R. R. Nair, R. Jalil, L. Britnell, B. D.Belle, E. W. Hill, K. S. Novoselov, K. Watanabe, T. Taniguchi,A. K. Geim, and P. Blake. Small 7, 465 (2011). T. Uwanno, Y. Hattori, T. Taniguchi, K. Watanabe, and K. Na-gashio. 2D Materials 2, 041002 (2015). W. Strupinski, K. Grodecki, A. Wysmolek, R. Stepniewski, T.Szkopek, P. E. Gaskell, A. Grneis, D. Haberer, R. Bozek, J.Krupka, and J. M. Baranowski. Nano Lett. 11, 1786 (2011). J. W. Suk, A. Kitt, C. W. Magnuson, Y. Hao, S. Ahmed, J. An,A. K. Swan, B. B. Goldberg, and R. S. Ruoff. ACS Nano 5, 6916(2011). N. David and K. Petr. Open Phys. 10, 181-188 (2012). S. Reich, a. Ferrari, R. Arenal, a. Loiseau, I. Bello, and J. Robert-son, Phys. Rev. B 71, 1 (2005). P. Trautman, K. Pakula, R. Bozek, and J. M. Baranowski, Appl.Phys. Lett. 83, 3510 (2003). G. Kowalski, S. P. Collins, and M. Moore J. App. Phys. 87, 3663(2000). P. Trautman and J. M. Baranowski, Phys. Rev. Lett. 69, 664(1992). M. Leszcznski, T. Suski, and G. Kowalski, Semicond. Sci. Tech-nol. 6, 59 (1991). B. Huang and H. Lee, Phys. Rev. B 86, 245406 (2012). P. Boguslawski and J. Bernholc, Phys. Rev. B 56, 9496 (1997). M. Noor-A-Alam, H. J. Kim, and Y.-H. Shin, Phys. Chem.Chem. Phys. 16, 6575 (2014). L. Ju, J. Velasco, E. Huang, S. Kahn, C. Nosiglia, H.-Z. Tsai,W. Yang, T. Taniguchi, K. Watanabe, Y. Zhang, G. Zhang, M.Crommie, A. Zettl, and F. Wang. Nat. Nanotech. 9, 348 (2014). J. M. Urban, P. Dbrowski, J. Binder, M. Kopciuszyski, A. Wys-moek, Z. Klusek, M. Jaochowski, W. Strupiski, and J. M. Bara-nowski J. Appl. Phys. 115, 233504 (2014). J. E. Lee, G. Ahn, J. Shim, Y. S. Lee, and S. Ryu, Nat. Comm.3, 1024 (2012).37