<?xml version="1.0" encoding="UTF-8"?>
<TEI xmlns="http://www.tei-c.org/ns/1.0" xml:lang="en">
  <teiHeader>
    <fileDesc>
      <titleStmt>
        <title>High-quality sandwiched black phosphorus heterostructure and its quantum oscillations</title>
        <author>Xiaolong Chen</author>
        <author>Yingying Wu</author>
        <author>Zefei Wu</author>
        <author>Yu Han</author>
        <author>Shuigang Xu</author>
        <author>Lin Wang</author>
        <author>Weiguang Ye</author>
        <author>Tianyi Han</author>
        <author>Yuheng He</author>
        <author>Yuan Cai</author>
        <author>Ning Wang</author>
      </titleStmt>
      <publicationStmt>
        <p>Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the
           staged copy named in sourceDesc and is not altered; the tags are held separately in
           standOff so that overlapping and nested spans are preserved.</p>
      </publicationStmt>
      <sourceDesc>
        <bibl>
          <title>High-quality sandwiched black phosphorus heterostructure and its quantum oscillations</title>
          <date>2015</date>
          <publisher>Nature Communications</publisher>
          <idno type="DOI">10.1038/ncomms8315</idno>
          
          <idno type="record">chen2015-sandwiched-black-phosphorus-quantum-oscillations</idno>
          <note type="staged-file">chen2015-sandwiched-black-phosphorus-quantum-oscillations.txt, the project's staged copy</note>
          <note type="text-sha1">e1bca6b0a84254a2b216e82c3797de373b65992a</note>
          <note type="rights">cc-by</note>
        </bibl>
      </sourceDesc>
    </fileDesc>
    <profileDesc>
      <textClass><keywords><term>physics</term></keywords></textClass>
      <particDesc>
        <person xml:id="paper"><p>the paper's own words</p></person>
        <person xml:id="tagger"><p>the tag layer; a sentence written by the tagger about the
          paper, always anchored to a quotation but not itself a quotation</p></person>
      </particDesc>
    </profileDesc>
    <revisionDesc>
      <change when="2026-10-06"
              who="build_paper_edition.py">50 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them</change>
    </revisionDesc>
  </teiHeader>
  <standOff>
    <spanGrp type="tags">
      <span xml:id="t0" from="#a3" to="#a4" type="object_of_study" ana="#object_of_study" resp="#paper">few-layer black phosphorus encapsulated between hexagonal boron nitride layers</span>
      <span xml:id="t1" from="#a1" to="#a2" type="research_question" ana="#research_question" resp="#paper">how to overcome degradation and charge traps that limit black phosphorus mobility</span>
      <span xml:id="t2" from="#a5" to="#a6" type="contribution" ana="#contribution" resp="#paper">a stable BN–BP–BN heterostructure with high mobility under ambient conditions</span>
      <span xml:id="t3" from="#a3" to="#a4" type="claim_type" ana="#claim_type" resp="#paper">method</span>
      <span xml:id="t4" from="#a17" to="#a18" type="phenomenon" ana="#phenomenon" resp="#paper">high hole mobility and Shubnikov-de Haas oscillations in encapsulated black phosphorus</span>
      <span xml:id="t5" from="#a19" to="#a20" type="method" ana="#method" resp="#paper">experiment<term>experiment</term></span>
      <span xml:id="t6" from="#a13" to="#a14" type="mechanisms" ana="#mechanisms[0]" resp="#tagger" cert="0.85">BN layers hide BP from ambient air and stop its reaction with O2-saturated water.<term>encapsulation</term></span>
      <span xml:id="t7" from="#a33" to="#a34" type="mechanisms" ana="#mechanisms[1]" resp="#tagger" cert="0.85">Ultra-clean interfaces and high-temperature annealing suppress charge trap states.<term>clean interfaces and annealing</term></span>
      <span xml:id="t8" from="#a43" to="#a44" type="mechanisms" ana="#mechanisms[2]" resp="#tagger" cert="0.8">A magnetic field breaks time-reversal symmetry, suppressing backscattering and lowering resistance.<term>weak localization</term></span>
      <span xml:id="t9" from="#a25" to="#a26" type="mechanisms" ana="#mechanisms[3]" resp="#tagger" cert="0.7">Transport along the fast X-direction gives higher mobility.<term>crystal direction</term></span>
      <span xml:id="t10" from="#a39" to="#a40" type="limitations" ana="#limitations[0]" resp="#tagger" cert="0.8">Vacancies and transfer-induced strain may still limit mobility, especially at low temperature.</span>
      <span xml:id="t11" from="#a29" to="#a30" type="limitations" ana="#limitations[1]" resp="#tagger" cert="0.75">Mobility remains below bulk BP values.</span>
      <span xml:id="t12" from="#a53" to="#a54" type="limitations" ana="#limitations[2]" resp="#tagger" cert="0.75">No SdH oscillations appear along the Y-direction up to 8 T.</span>
      <span xml:id="t13" from="#a9" to="#a10" type="concepts" ana="#concepts[0]" resp="#paper" cert="0.9">BP<term>BP</term></span>
      <span xml:id="t14" from="#a1" to="#a2" type="concepts" ana="#concepts[1]" resp="#paper" cert="0.7">black phosphorus<term>black phosphorus</term></span>
      <span xml:id="t15" from="#a15" to="#a16" type="concepts" ana="#concepts[2]" resp="#paper" cert="0.95">BN<term>BN</term></span>
      <span xml:id="t16" from="#a15" to="#a16" type="concepts" ana="#concepts[3]" resp="#paper" cert="0.7">hexagonal boron nitride<term>hexagonal boron nitride</term></span>
      <span xml:id="t17" from="#a9" to="#a10" type="concepts" ana="#concepts[4]" resp="#paper" cert="0.95">FETs<term>FETs</term></span>
      <span xml:id="t18" from="#a9" to="#a10" type="concepts" ana="#concepts[5]" resp="#paper" cert="0.7">field-effect transistors<term>field-effect transistors</term></span>
      <span xml:id="t19" from="#a17" to="#a18" type="concepts" ana="#concepts[6]" resp="#paper" cert="0.95">SdH<term>SdH</term></span>
      <span xml:id="t20" from="#a17" to="#a18" type="concepts" ana="#concepts[7]" resp="#paper" cert="0.7">Shubnikov-de Haas<term>Shubnikov-de Haas</term></span>
      <span xml:id="t21" from="#a49" to="#a50" type="concepts" ana="#concepts[8]" resp="#paper" cert="0.95">LL<term>LL</term></span>
      <span xml:id="t22" from="#a49" to="#a50" type="concepts" ana="#concepts[9]" resp="#paper" cert="0.7">Landau level<term>Landau level</term></span>
      <span xml:id="t23" from="#a23" to="#a24" type="concepts" ana="#concepts[10]" resp="#paper" cert="0.7">charge trap density<term>charge trap density</term></span>
      <span xml:id="t24" from="#a23" to="#a24" type="concepts" ana="#concepts[11]" resp="#paper" cert="0.7">annealing<term>annealing</term></span>
      <span xml:id="t25" from="#a23" to="#a24" type="concepts" ana="#concepts[12]" resp="#paper" cert="0.7">hysteresis<term>hysteresis</term></span>
      <span xml:id="t26" from="#a27" to="#a28" type="concepts" ana="#concepts[13]" resp="#paper" cert="0.7">mobility<term>mobility</term></span>
      <span xml:id="t27" from="#a37" to="#a38" type="concepts" ana="#concepts[14]" resp="#paper" cert="0.7">phonon scattering<term>phonon scattering</term></span>
      <span xml:id="t28" from="#a41" to="#a42" type="concepts" ana="#concepts[15]" resp="#paper" cert="0.7">weak localization<term>weak localization</term></span>
      <span xml:id="t29" from="#a41" to="#a42" type="concepts" ana="#concepts[16]" resp="#paper" cert="0.7">negative magneto-resistance<term>negative magneto-resistance</term></span>
      <span xml:id="t30" from="#a45" to="#a46" type="concepts" ana="#concepts[17]" resp="#paper" cert="0.7">Berry phase<term>Berry phase</term></span>
      <span xml:id="t31" from="#a51" to="#a52" type="concepts" ana="#concepts[18]" resp="#paper" cert="0.7">cyclotron effective mass<term>cyclotron effective mass</term></span>
      <span xml:id="t32" from="#a7" to="#a8" type="concepts" ana="#concepts[19]" resp="#paper" cert="0.7">Graphene<term>Graphene</term></span>
      <span xml:id="t33" from="#a7" to="#a8" type="concepts" ana="#concepts[20]" resp="#paper" cert="0.7">bandgap<term>bandgap</term></span>
      <span xml:id="t34" from="#a31" to="#a32" type="senses" ana="#senses[0]" resp="#tagger" cert="0.7">quality - how free a BP sample is of charge traps, defects and degradation, as seen in mobility</span>
      <span xml:id="t35" from="#a15" to="#a16" type="concept_edges" ana="#concept_edges[0]" resp="#tagger" cert="0.95">BN —same-as→ hexagonal boron nitride<note>the relation itself is separately evidenced · strength of the relation: 0.95</note><term>BN</term><term>hexagonal boron nitride</term></span>
      <span xml:id="t36" from="#a9" to="#a10" type="concept_edges" ana="#concept_edges[1]" resp="#tagger" cert="0.95">FETs —same-as→ field-effect transistors<note>the relation itself is separately evidenced · strength of the relation: 0.95</note><term>FETs</term><term>field-effect transistors</term></span>
      <span xml:id="t37" from="#a17" to="#a18" type="concept_edges" ana="#concept_edges[2]" resp="#tagger" cert="0.95">SdH —same-as→ Shubnikov-de Haas<note>the relation itself is separately evidenced · strength of the relation: 0.95</note><term>SdH</term><term>Shubnikov-de Haas</term></span>
      <span xml:id="t38" from="#a49" to="#a50" type="concept_edges" ana="#concept_edges[3]" resp="#tagger" cert="0.95">LL —same-as→ Landau level<note>the relation itself is separately evidenced · strength of the relation: 0.95</note><term>LL</term><term>Landau level</term></span>
      <span xml:id="t39" from="#a23" to="#a24" type="concept_edges" ana="#concept_edges[4]" resp="#tagger" cert="0.85">annealing —causes→ charge trap density<note>direction on the measure: reduces · the relation itself is separately evidenced · strength of the relation: 0.85 · the relation stated: can significantly reduce</note><term>annealing</term><term>charge trap density</term></span>
      <span xml:id="t40" from="#a35" to="#a36" type="content_tags" ana="#content_tags.findings[0]" resp="#tagger" cert="0.9">Mobility and on–off ratios show no degradation after a week in ambient conditions.</span>
      <span xml:id="t41" from="#a21" to="#a22" type="content_tags" ana="#content_tags.findings[1]" resp="#tagger" cert="0.85">Without BN protection, few-layer BP breaks down at 350 °C.</span>
      <span xml:id="t42" from="#a49" to="#a50" type="content_tags" ana="#content_tags.findings[2]" resp="#tagger" cert="0.85">A linear Landau fit gives a Berry phase of zero.</span>
      <span xml:id="t43" from="#a47" to="#a48" type="content_tags" ana="#content_tags.findings[3]" resp="#tagger" cert="0.85">The carriers are confirmed as two-dimensional by the 1/cos θ dependence.</span>
      <span xml:id="t44" from="#a27" to="#a28" type="content_tags" ana="#content_tags.findings[4]" resp="#tagger" cert="0.8">Mobility along X can be twice that along Y.</span>
      <span xml:id="t45" from="#a15" to="#a16" type="content_tags" ana="#content_tags.definitions[0]" resp="#tagger" cert="0.85">few-layer phosphorene encapsulated by atomically thin hexagonal boron nitride<term>BN–BP–BN heterostructure</term></span>
      <span xml:id="t46" from="#a17" to="#a18" type="content_tags" ana="#content_tags.quantities[0]" resp="#tagger" cert="0.9">hole mobility at cryogenic temperature</span>
      <span xml:id="t47" from="#a51" to="#a52" type="content_tags" ana="#content_tags.quantities[1]" resp="#tagger" cert="0.9">cyclotron effective mass at −60 V gate</span>
      <span xml:id="t48" from="#a11" to="#a12" type="content_tags" ana="#content_tags.quantities[2]" resp="#tagger" cert="0.85">earlier room-temperature mobility limit of few-layer phosphorene</span>
      <span xml:id="t49" from="#a27" to="#a28" type="content_tags" ana="#content_tags.variables[0]" resp="#tagger" cert="0.8">mobility<term>mobility</term></span>
    </spanGrp>
  </standOff>
  <text>
    <body>
        <p><anchor xml:id="a0"/> High-quality sandwiched black phosphorus heterostructure and its quantum oscillations</p>
        <p>Abstract</p>
        <p>Two-dimensional materials such as graphene and transition metal dichalcogenides have attracted great attention because of their rich physics and potential applications in next-generation nanoelectronic devices. The family of two-dimensional materials was recently joined by atomically thin black phosphorus which possesses high theoretical mobility and tunable bandgap structure. <anchor xml:id="a1"/>However, degradation of properties under atmospheric conditions and high-density charge traps in black phosphorus have largely limited its actual mobility thus hindering its future applications.<anchor xml:id="a2"/> <anchor xml:id="a3"/>Here, we report the fabrication of stable sandwiched heterostructures by encapsulating atomically thin black phosphorus between hexagonal boron nitride layers to realize ultra-clean interfaces<anchor xml:id="a4"/> that allow a high field-effect mobility of ∼1,350 cm 2 V −1 s −1 at room temperature and on–off ratios exceeding 10 5 . At low temperatures, the mobility even reaches ∼2,700 cm 2 V −1 s −1 and quantum oscillations in black phosphorus two-dimensional hole gas are observed at low magnetic fields. <anchor xml:id="a5"/>Importantly, the sandwiched heterostructures ensure that the quality of black phosphorus remains high under ambient conditions.<anchor xml:id="a6"/></p>
        <p>Two-dimensional (2D) materials with both high mobility and high on–off ratios are in demand for next-generation nanodevices. <anchor xml:id="a7"/>Graphene 1 , 2 , being the most widely studied material, has shown rich physics and high mobility, but the absence of a bandgap 1 , 2 limits its applications.<anchor xml:id="a8"/> Analogous to 2D transition metal dichalcogenide structures 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , atomically thin black phosphorus (BP) 12 , 13 , 14 , 15 has recently evoked interest due to its high theoretical mobility 16 , tunable direct bandgap 17 , 18 , 19 , 20 and ambipolarity 12 , 13 , 14 , 15 —unique properties that make it suitable for application in electronic and optoelectronic devices 21 , 22 , 23 . <anchor xml:id="a9"/>Recently, field-effect transistors (FETs) 12 , 13 , 14 , 15 , 24 , 25 , 26 , electrically tunable PN junctions 22 , radio-frequency transistors 27 and heterojunctions 28 , 29 based on few-layer BP have also been demonstrated.<anchor xml:id="a10"/> <anchor xml:id="a11"/>However, the room temperature mobility of few-layer phosphorene reported so far is limited to 400 cm 2 V −1 s −1 due to the presence of high-density charge traps, defects and phonon scattering 12 , 13 , 14 , 15 .<anchor xml:id="a12"/> <anchor xml:id="a13"/>The quality degradation of BP under atmospheric conditions is mainly due to its reaction with O 2 -saturated H 2 O 30 , 31 .<anchor xml:id="a14"/></p>
        <p>Here we demonstrate a high field-effect mobility of ∼1,350 cm 2 V −1 s −1 at room temperature and high on–off ratios exceeding 10 5 <anchor xml:id="a15"/>in few-layer phosphorene encapsulated by atomically thin hexagonal boron nitride (BN), which forms a stable BN–BP–BN heterostructure.<anchor xml:id="a16"/> A room temperature mobility of ∼1,000 cm 2 V −1 s −1 is rare in 2D semiconductor electron gas systems prepared by mechanical exfoliation techniques. In addition, the BN layers further avoids quality degradation of BP when exposed to atmosphere. <anchor xml:id="a17"/>At cryogenic temperatures, the mobility even reaches ∼2,700 cm 2 V −1 s −1 and Shubnikov-de Haas (SdH) oscillations are observed in BP at low magnetic fields (∼6 T).<anchor xml:id="a18"/> Our fabrication and treatment techniques for BN–BP–BN heterostructures open up a way to achieve high-performance 2D semiconductors with greatly improved room temperature mobility, on–off ratios and stability under ambient conditions for practical applications in high-speed electronic and optoelectronic devices.</p>
        <p>Results</p>
        <p>Fabrication of the sandwiched BP heterostructure devices</p>
        <p><anchor xml:id="a19"/>To achieve both high mobility and stability of BP FETs under atmospheric conditions, the BN–BP–BN configuration and high-temperature annealing are the two key factors.<anchor xml:id="a20"/> The ultra-clean BN–BP interfaces are ensured by adopting the polymer-free van der Waals transfer technique 32 as shown in Fig. 1a . The few-layer BP mechanically exfoliated on a 300-nm SiO 2 /Si substrate was first picked up by a thin BN flake (6–20 nm thick). Then the BN–BP sample was transferred to a BN flake supported on a SiO 2 /Si substrate to form the BN–BP–BN heterostructure. The atomically thin BP was completely encapsulated by two BN layers, thus allowing us to anneal the sample at temperatures up to 500 °C in an argon atmosphere to further improve the sample quality. <anchor xml:id="a21"/>Without the BN protective layers, few-layer BP breaks down easily at 350 °C (see Supplementary Note 1 and Supplementary Fig. 1 ).<anchor xml:id="a22"/> <anchor xml:id="a23"/>In addition, the annealing process can significantly reduce the charge trap density in BP as no hysteresis effect is observed at room temperature<anchor xml:id="a24"/> ( Fig. 2g and Supplementary Fig. 2 ).</p>
        <p>BN–BP–BN heterostructure device. ( a ) Schematic of the BN–BP–BN heterostructure device fabrication process. ( b , e ) Schematic ( b ) and optical image ( e ) of a BN–BP–BN Hall-bar device. Scale bar, 10 μm. ( c ) Atomic structure of monolayer BP. ( d ) The BN–BP–BN heterostructure after O 2 -plasma etching. The etched area is enclosed within the white line. The purple-dashed line denotes the lower BN layer.</p>
        <p>Mobility and stability of the BN–BP–BN heterostructure devices. ( a ) The I – V ds curves obtained at different gate voltages at 1.7 K. The inset is the optical image of Sample A with the following geometrical parameters: L 14 =16 μm, L 23 =10 μm and W =3 μm. ( b ) The resistivity determined from four-terminal (green line) and two-terminal configurations (orange line) at 1.7 K. ( c ) Variation of the contact resistivity. ( d ) The conductivity of Sample A measured at with a room temperature and 1.7 K. The inset shows the ambipolarity of the BP conductance. ( e , f ) Temperature dependence of the field-effect μ F (open dots) and Hall mobility μ h (solid dots) at V g =−70 V) of Sample A and Sample B (15 nm thick). The dashed lines serve as guidelines for the μ ∼ T − γ relation. ( g ) The room temperature conductivity showing no hysteresis in Sample A. ( h ) The mobility and on–off ratio of Sample A as a functions of ambient exposure time. No quality degradation is observed even after exposure for an entire week.</p>
        <p>Instead of using one-dimensional edge-contact 32 that has been proven effective for graphene-encapsulated structures (but less so for 2D semiconductors), we developed the area contact technique for contacting the encapsulated 2D semiconductors. Beginning with the BN–BP–BN heterostructure, a hard mask is defined by the standard electron-beam lithography technique using ZEP-520 resist (detailed process is shown in Supplementary Note 2 and Supplementary Fig. 3 ). Since the O 2 -plasma etching rates for BN and BP are different, the BN layers can be quickly etched while the BP layer still survive as shown in Fig. 1d . Lastly, Cr/Au (2 nm/60 nm) electrodes are deposited using electron-beam lithography technique. Figure 1b,e show the schematic and optical image of a BN–BP–BN Hall-bar device respectively. Note that no further annealing process is needed after deposition of the electrodes.</p>
        <p>Mobility and stability of the sandwiched BP devices</p>
        <p>Figure 2a shows the I – V ds characteristic of an 8-nm-thick BN–BP–BN sample (Sample A). The linear I – V ds characteristic indicates that Ohmic contacts have been achieved on the hole side. By comparing the resistivity directly measured using the two-terminal configuration ( R 14 W / L 14 ) with that extracted using the four-terminal configuration ( R 23 W / L 23 ), the contact resistivity is obtained ( Fig. 2b ). As shown in Fig. 2c , the contact resistivity measured at high gate voltages is about several when the sample temperature is sufficiently low and thermal activation effects are suppressed. The contact resistivity decreases with decreasing temperature when the gate voltage V g &lt;−68 V, further verifying the good contacts between BP and Cr/Au electrodes on the hole side.</p>
        <p>The BN–BP–BN heterostructure exhibits high quality as is first confirmed by examining the four-terminal field-effect (FET) mobility at different temperatures, where σ is the conductivity and C g is the gate capacitance. Figure 2d shows the conductivity as a function of gate voltage for Sample A. Similar to the results reported previously, both hole and electron conductance can be achieved (see the inset in Fig. 2d ). Here we mainly focus on the hole conductance due to its higher mobility. The gate capacitance C g is calculated to be 1.1 F cm −2 based on the thickness of SiO 2 (300 nm) and the lower BN layer (∼15 nm as determined by atomic force microscopy). With this information, we extracted the hole μ F from the linear part of conductance at different temperatures ( Fig. 2e,f ). As shown in Fig. 2d , we achieved a high FET mobility μ F ∼1,350 cm 2 V −1 s −1 and Hall mobility μ H ∼790 cm 2 V −1 s −1 for Sample A at room temperature. Another four BN–BP–BN heterostructures also exhibit high mobility (∼800 cm 2 V −1 s −1 ) at room temperature (summarized in Supplementary Table 1 ). <anchor xml:id="a25"/>One possible reason for the high mobility in our BP samples is because their transport directions are nearly along the fast X-direction<anchor xml:id="a26"/> (see Fig. 1c ) as confirmed by angle-dependent polarized Raman spectroscopy (see Supplementary Note 3 and Supplementary Fig. 4 ). <anchor xml:id="a27"/>The mobility along the X-direction can be two times larger than that along the Y-direction 13 , 15 .<anchor xml:id="a28"/> The transport-direction of Sample A is parallel to the X-direction. <anchor xml:id="a29"/>Although the mobility we achieved is still smaller than the values measured from bulk BP (∼2,000 cm 2 V −1 s −1 ) 33 , it is much larger than that observed in atomically thin MoS 2 (refs 3 , 4 , 10 ) and few-layer BP supported on SiO 2 (refs 12 , 13 , 14 , 15 ).<anchor xml:id="a30"/> Theoretical calculations for few-layer BP 16 indicate that a hole mobility over 4,000 cm 2 V −1 s −1 is achievable along the X-direction at room temperature. <anchor xml:id="a31"/>Hence, the quality improvement is still the critical limiting factor for the mobility of few-layer BP.<anchor xml:id="a32"/></p>
        <p>The high quality of the BN–BP–BN heterostructure is further confirmed by the observation of a high on–off ratio exceeding 10 5 and insignificant hysteresis at room temperature ( Fig. 2g ), while previous work 14 on BP supported on SiO 2 showed a pronounced hysteresis (Δ V g ∼80 V) depending on the sweep direction at room temperature, indicating the presence of a high density of charge traps. The small charge trap density in our BN–BP–BN heterostructure is due not only to the ultra-clean BN–BP interfaces but also to the high-temperature annealing (300–500 °C) of the encapsulated BP. As shown in Supplementary Fig. 2 , in a BN–BP–BN heterostructure device without annealing, we observed a hysteresis Δ V g ∼10 V, which is much smaller than that observed in BP supported on SiO 2 . Supplementary Note 4 and Supplementary Fig. 5 demonstrate the mobility changes of Sample D before and after annealing. The mobility of this sample measured at cryogenic and room temperatures indeed increases significantly. <anchor xml:id="a33"/>These results verified that high-temperature annealing can effectively suppress charge trap states and further improve the mobility and on–off ratio of BP devices.<anchor xml:id="a34"/></p>
        <p>The stability of the electrical performance of the BN–BP–BN heterostructure is examined under ambient conditions (with humidity around 60%). <anchor xml:id="a35"/>As shown in Fig. 2h , the mobility and on–off ratios show no degradation even after exposure to ambient conditions for a whole week.<anchor xml:id="a36"/> Obviously, the BN–BP–BN heterostructure completely hides the BP layer from ambient conditions and prevents it from reacting with O 2 -saturated H 2 O. With its high room temperature mobility, on–off ratios and stability, the BN–BP–BN heterostructure is a promising candidate for practical applications in electronic and optoelectronic devices.</p>
        <p>At 1.7 K, the mobility of Sample A reaches ∼2,700 cm 2 V −1 s −1 ( Fig. 2d ) and the Hall mobility is ∼1,500 cm 2 V −1 s −1 at large gate voltages, where n H is the carrier density from Hall measurement, which is equal to the value extracted from the gate capacitance model n = C g ( V g – V th )/ e (see Supplementary Note 5 and Supplementary Fig. 6 ). The on–off ratios exceed 10 8 at 1.7 K. The temperature dependence of the FET and Hall mobility of the BN–BP–BN heterostructures ( Fig. 2e,f ) shows a similar trend as reported previously for few-layer phosphorene 12 and monolayer MoS 2 (ref. 4 ). <anchor xml:id="a37"/>The mobility saturates at low temperatures ( T &lt;80 K), but follows the expression μ ∼ T − γ at higher temperatures ( T &gt;80 K) due to phonon scattering.<anchor xml:id="a38"/> The extracted value γ is ∼0.57 for Sample A and ∼0.69 for Sample B (15 nm thick). These γ values are consistent with the ones observed in monolayer MoS 2 covered with high-κ dielectric 4 and few-layer phosphorene 12 , but are smaller than those found in bulk BP 34 and other 2D materials 35 . Similar to the results reported previously 12 , 13 , 15 , the values of the Hall mobility measured from our samples are always smaller than that of their FET mobility ( Supplementary Table 1 ). <anchor xml:id="a39"/>We believe that although charge traps are effectively suppressed in our BN–BP–BN heterostructures through annealing, other types of defects/impurities, such as vacancies and strain induced during the transfer process may still exist.<anchor xml:id="a40"/> These defects limit the mobility of few-layer BP particularly at low temperatures.</p>
        <p>Quantum oscillations in the sandwiched BP devices</p>
        <p>The quantum oscillations observed at magnetic fields B &gt;6 T also attest to the high quality of BN–BP–BN heterostructures. <anchor xml:id="a41"/>Resistance at different gate voltages shows a negative magneto-resistance ( Fig. 3a ), indicating a weak localization effect 36 previously observed in 2D electron gas systems containing disorder.<anchor xml:id="a42"/> <anchor xml:id="a43"/>In this kind of 2D electron gas systems, the time-reversal symmetry of the closed electron path is broken when a magnetic field is applied and the back scattering of electrons is suppressed resulting in a negative correction to the resistance.<anchor xml:id="a44"/> When B &gt;6 T (corresponding to a mobility of ∼1/ B near 1,670 cm 2 V −1 s −1 ), we start to see SdH oscillations at higher gate voltages where the Hall mobility becomes sufficiently high ( Fig. 3a ). The oscillation features are seen clearly in the form of d R /d B as shown in Fig. 3b (also see the data of Sample B in Supplementary Note 6 and Supplementary Figs 7 and 8 ).</p>
        <p>SdH oscillations in the 8-nm-thick BN–BP–BN heterostructure. ( a ) Resistance ( R ) plotted as a function of the magnetic field at gate voltages of −50 V (blue dots), −60 V (green dots) and −70 V (purple dots). ( b ) d R /d B plotted as a function of 1/ B . The dashed lines indicate the oscillation period (increasing with gate voltages) of 1/ B F . ( c ) Δ R plotted as a function of 1/ B at V g =−60 V for different temperatures. ( d ) Fitting of the experimental results (dots) using the Lifshitz–Kosevich formula (solid line). The inset shows the fitting results of the cyclotron mass. ( e ) Landau diagram at different gate voltages that shows a Berry phase β =0. ( f ) The carrier densities determined from LL fitting (orange dots) and gate capacitance (purple line).</p>
        <p>The period of SdH oscillations 37 , 38 in 2D electron gas is 1/ B F and the longitudinal resistance is given by Δ R = R ( B , T )cos[2 π ( B F / B +1/2+ β )], where R ( B , T ) is the amplitude and β (0≤ β ≤1) is the Berry phase. <anchor xml:id="a45"/>β is known to be 0 in conventional 2D electron gas systems and 1/2 in graphene 1 , 2 .<anchor xml:id="a46"/> To examine the 2D nature of hole gas in few-layer BP, we plot d R /d B as a function of 1/ B for different gate voltages. As shown in Fig. 3b , the oscillations at different gate voltages are periodic but exhibit different frequencies. A smaller period 1/ B F is observed at higher gate voltages where the carrier density n is larger, which is consistent with the theoretical oscillation period 37 expressed as 1/ B F =2/Φ 0 n , where Φ 0 =4.14 × 10 −15 Tm 2 is the flux quantum. For example, at V g =−50 V the oscillation period is ∼0.0138 T −1 , corresponding to a carrier density of ∼3.5 × 10 12 cm −2 , which is in excellent agreement with the value obtained from the gate capacitance model n = C g ( V g − V th )/ e ( Fig. 3f ). To further reveal the 2D nature of the charge carriers in few-layer BP, we carried out experiments to investigate the angular dependence of the SdH oscillations. <anchor xml:id="a47"/>The linear dependence between the magnetic field B F and 1/cos θ confirms that the charge carriers in few-layer BP are in 2D nature<anchor xml:id="a48"/>, where θ is the tilting angle between the applied magnetic fields and the normal direction of the substrate plane (see Supplementary Note 7 and Supplementary Fig. 9 ). <anchor xml:id="a49"/>The Landau level (LL) index N is plotted and the linear fit yields a Berry phase β =0 in our BP samples ( Fig. 3e ).<anchor xml:id="a50"/> A smaller LL index would be achieved if magnetic fields could be further increased.</p>
        <p>The cyclotron effective mass m c of hole carriers in few-layer BP can be obtained by fitting the temperature dependence of resistivity Δ R (see Fig. 3c ) using the Lifshitz–Kosevich formula A T ∝ T /sinh (2 π 2 k B Tm c / ℏ eB ), where k B is the Boltzmann constant. We fit Δ R at high magnetic fields to gain a better resolution as shown in Fig. 3d . <anchor xml:id="a51"/>The fitting results produce a cyclotron effective mass 0.27±0.02 m e at V g =−60 V.<anchor xml:id="a52"/> Owing to structural anisotropy, the carrier effective mass in few-layer BP is also highly anisotropic. Previous theoretical analyses 16 show that the effective masses along the X- and Y- directions are m x ∼0.14 m e and m y ∼0.89 m e respectively. The measured cyclotron effective mass in our BP samples is ∼0.27 m e , which is reasonable as compared with the theoretical value for few-layer BP 16 .</p>
        <p>Finally, we designed a special sample (Sample C, 10 nm thick) to demonstrate the effects of transport anisotropy on the SdH oscillations in BP. First, this sample was cut along the X-direction ( Fig. 4a ). The Hall mobility was ∼1,481 cm 2 V −1 s −1 at 1.7 K (∼716 cm 2 V −1 s −1 at room temperature). The SdH oscillations are clearly visible in Fig. 4b when the magnetic field is higher than 6 T. Then, the sample was reshaped along the Y-direction (see the inset in Fig. 4a ). The Hall mobility changed to ∼820 cm 2 V −1 s −1 at 1.7 K (∼365 cm 2 V −1 s −1 at room temperature). This experimental results are consistent with the conclusions in previous reports 15 , 33 . <anchor xml:id="a53"/>However, the magneto-resistivity in this reshaped sample (along the Y-direction) did not show any signature of SdH oscillations with magnetic fields up to 8 T.<anchor xml:id="a54"/> A much stronger magnetic field is needed to observe SdH oscillations in the sample.</p>
        <p>SdH oscillations in the BN–BP–BN heterostructure samples along the X- and Y- directions. ( a ) The conductivity of Sample C along the X- and Y- directions at 1.7 K and 300 K. The insets are optical images of Sample C before and after reshaping. ( b ) SdH oscillations measured at V g =−60 V from the samples along the X- and Y- directions.</p>
        <p>Discussion</p>
        <p>In summary, we have demonstrated high-quality BN–BP–BN heterostructures achieved by encapsulating atomically thin BP between BN layers followed by annealing at high temperature. The BN–BP–BN heterostructure protects BP from property degradation and allows us to investigate the intrinsic properties of BP. The BN–BP–BN samples show excellent stability in the atmospheric environment with a high mobility ∼1,350 cm 2 V −1 s −1 and a large on–off ratio exceeding 10 5 . Quantum oscillations and zero Berry phase are observed in BP hole gas at a magnetic field of 6 T at cryogenic temperatures. The ultra-clean interfaces realized by our fabrication process can effectively suppress charge trap states. They shed important light on how to improve the quality of atomically thin BP for practical applications in BP-based nanoelectronic devices.</p>
        <p>Methods</p>
        <p>Sample preparation and transport measurements</p>
        <p>Bulk phosphorus and BN crystals (Polartherm grade PT110) were purchased from Smart-elements and Momentive, respectively. All atomically thin flakes were mechanically exfoliated by the scotch-tape microcleavage method. The thickness of BN and BP flakes was determined using an atomic force microscope (Veeco-Innova). The BN–BP–BN heterostructure was fabricated using the polymer-free van der Waals transfer technique 32 . The encapsulated BP was annealed in an argon atmosphere at 300–500 °C for 8 h before the upper BN layer was etched. No further annealing process was needed after deposition of the electrodes. Electrical measurements were performed using lock-in techniques in a cryogenic system (1.7–300 K and magnetic field up to 8 T).</p>
        <p>Additional information</p>
        <p>How to cite this article: Chen, X. et al. High quality sandwiched black phosphorus heterostructure and its quantum oscillations. Nat. Commun. 6:7315 doi: 10.1038/ncomms8315 (2015).</p>
        <p>Supplementary Material</p>
        <p>Supplementary Figures 1-9, Supplementary Tables 1, Supplementary Notes 1-7 and Supplementary References.</p>
        <p>Acknowledgments</p>
        <p>Financial support from the Research Grants Council of Hong Kong (Project Nos HKU9/CRF/13G, 604112, HKUST9/CRF/08 and N_HKUST613/12) and technical support of the Raith-HKUST Nanotechnology Laboratory for the electron-beam lithography facility at MCPF (Project No. SEG_HKUST08) are hereby acknowledged.</p>
        <p>Footnotes</p>
        <p>References</p>
        <p>Associated Data</p>
        <p>Supplementary Materials</p>
        <p>Supplementary Figures 1-9, Supplementary Tables 1, Supplementary Notes 1-7 and Supplementary References.<anchor xml:id="a55"/></p>
    </body>
  </text>
</TEI>
