sensors
Article Sensitive Planar Microwave Diode on the Base of Ternary AlxGa1-xAs Semiconductor Compound
Maksimas Anbinderis 1,2, Steponas Ašmontas 1, Aurimas Cerškusˇ 1,2, Jonas Gradauskas 1,2 , Andžej Luˇcun 1, Aldis Šilenas˙ 1 and Algirdas Sužiedelis˙ 1,*
1 Center for Physical Sciences and Technology, Savanoriu˛Ave. 231, 02300 Vilnius, Lithuania; [email protected] (M.A.); [email protected] (S.A.); [email protected] (A.C.);ˇ [email protected] (J.G.); [email protected] (A.L.); [email protected] (A.Š.) 2 Department of Physics, Vilnius Gediminas Technical University, Sauletekio˙ Ave. 11, 10223 Vilnius, Lithuania * Correspondence: [email protected]; Tel.: +370-5-2124539
Abstract: The article presents the results of experimental studies of the dc and high-frequency electri-
cal characteristics of planar microwave diodes that are fabricated on the base of the n-AlxGa1-xAs layer (x = 0, 0.15 or 0.3), epitaxially grown on a semi-insulating GaAs substrate. The diodes can serve as reliable and inexpensive sensors of microwave radiation in the millimeter wavelength range; they sense electromagnetic radiation directly, without any external bias voltage at room temperature. The investigation revealed a strong dependence of the detection properties of the microwave diodes on AlAs mole fraction x: in the Ka microwave frequency range, the median value of voltage responsivity is several volts per watt in the case of GaAs-based diodes (x = 0), and it substantially increases, reaching hundreds of volts per watt at higher x values. Also, a model enabling us to forecast the responsivity of the sensor in other frequency ranges is proposed.
Citation: Anbinderis, M.; Ašmontas, S.; Cerškus,ˇ A.; Gradauskas, J.; Luˇcun, Keywords: planar diode; microwave sensor; voltage responsivity; aluminum gallium arsenide A.; Šilenas,˙ A.; Sužiedelis,˙ A. Sensitive Planar Microwave Diode on the Base of Ternary AlxGa1-xAs Semiconductor Compound. Sensors 2021, 21, 4487. 1. Introduction https://doi.org/10.3390/s21134487 Electromagnetic radiation in the millimeter wavelength range attracts the attention of scientists and engineers due to its numerous possible applications in modern fields of Academic Editor: Francesco technology, such as the imaging of concealed objects [1–3], inspection of material homo- Della Corte geneity [4,5], and arranging broadband cellular communication networks [6,7]. Successive use of the millimeter waves in frequency-modulated continuous wave (FMCW) radar Received: 25 May 2021 technology, in conjunction with small-sized MIMO (multiple-input multiple-output) an- Accepted: 26 June 2021 Published: 30 June 2021 tenna arrays, gives possibility for the precise scanning of nearby situated various objects. Texas Instruments Inc. produces low-cost automotive and industrial millimeter wave
Publisher’s Note: MDPI stays neutral sensors for radaring either automotive vehicles or smaller industrial objects [8,9]. The with regard to jurisdictional claims in fast development of such technologies requires novel concepts of electromagnetic radia- published maps and institutional affil- tion sensors. Some millimeter wave sensors require bulky cryogenic cooling equipment iations. (high-Tc Josephson junction detector [10] or Y–Ba–Cu–O microbolometer [11]). Some of them are quite complex, such as a Golay cell, which, on the other hand, demonstrates its advantage of high-voltage responsivity at room temperature and the ability to sense elec- tromagnetic radiation in a wide spectral range, from ultraviolet light up to Ka microwave (MW) frequencies [12]. Backward tunnel diodes [13], Si CMOS field-effect transistors [14], Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. and a big variety of microwave diodes on the base of Schottky junction [15] are also known This article is an open access article as room-temperature sensors of the millimeter waves. The most popular of them are distributed under the terms and zero-biased low-barrier Schottky diodes. These are favorable in small-signal square law conditions of the Creative Commons and envelope detection [16,17], and thus are useful in the application of wireless electro- Attribution (CC BY) license (https:// magnetic communication. High-voltage responsivity, low-noise characteristics, operation creativecommons.org/licenses/by/ at room temperature, wide bandwidth, competent reliability, and good repeatability make 4.0/). the Schottky diodes unquestioned leaders in the square detection race. The Schottky
Sensors 2021, 21, 4487. https://doi.org/10.3390/s21134487 https://www.mdpi.com/journal/sensors Sensors 2021, 21, 4487 2 of 17
diodes have reached their commercial maturity, allowing the purchase of these diodes in the market [18]. However, the fundamental concept of a Schottky junction requires it to be formed on a semiconductor surface that makes the diode’s properties sensitive to the manufacturing conditions and vulnerable to the impact of environment. Therefore, scientists all around the world are looking for new concepts of electromagnetic radiation sensors having detection processes situated in the bulk of a device. More than fifty years ago, promising microwave diodes operating on the basis of hot carrier phenomena in semiconductors, under the action of a nonhomogeneous electric field, were developed [19]. Further improvements in these point-contact diodes turned into the planar bow-tie design of the diodes, successfully sensing electromagnetic radiation from the microwaves to the infrared, including the terahertz frequency range [20–22]. These bow-tie diodes were successfully used for heterodyne and spectroscopic terahertz imaging and sensing [22,23]. Low-voltage responsivity was the main drawback of the bow-tie diode; it was increased by means of partial gating of the active two-dimensional electron gas layer in the vicinity of the diode’s contacts [24]. Another kind of microwave diodes having the active region situated in the bulk of a device is the heterojunction diode. In the sense of increasing their voltage responsivity, the point-contact [25] and planar [26] heterojunction diodes revealed themselves as counterparts of the hot carrier microwave diodes that are worth attention. As for commercial application, the microwave sensors must be not only sensitive and fast, but also reliable and inexpensive. We have proposed a simple design of an inexpensive planar dual microwave diode, fabricated on the base of low-resistivity gallium arsenide (GaAs) substrate [27]. The diodes were suitable both to sense microwave continuous wave signals and to measure pulsed microwave power in the nanosecond time scale [28]. However, such fast diodes, which can detect pulses of nanosecond duration, suffered from low-voltage responsivity, while more sensitive ones had high electrical resistance in the MΩ range. Thus, the main issue of this paper is increasing the voltage responsivity of the microwave radiation sensor on the base of the planar dual microwave diode, and narrowing the spread of its electrical parameters by means of its fabrication on the base of epitaxial aluminum gallium arsenide n-AlxGa1-xAs layer, grown on a semi-insulating GaAs substrate. The paper describes a detailed procedure of the sample fabrication, as well as nam- ing the measurement technique used and the research methodology applied (Section2). Section3 presents the main results of the investigation and discussion concerning the following: (i) statistical data of three batches of planar microwave diodes; (ii) dependence of electrical parameters of the representative diodes on illumination and temperature; (iii) fitting experimental data of dc and high-frequency parameters of the sensors using phenomenological theory; and (iv) evaluation of the ability of the planar microwave diode to withstand the impact of high-power microwave radiation.
2. Samples and Measurement Technique
The active region of the microwave diode was epitaxial n-type AlxGa1-xAs layer of sub- micrometric thickness grown onto a semi-insulating GaAs (100) substrate (350 AGChP-I-26a (100) 0◦300, Ø26 mm, Plant of Pure Metals, Svitlovodsk, Ukraine) by means of liquid phase epitaxy technique in the quartz tube furnace “SDO-125/3” (Termotron, Bryansk, USSR). Epi- taxial layers were grown using the supercooled growth method in “wiping-less” horizontal graphite sliding boat. Initial supercooling was performed at 6 ◦C below the melting point of Ga–GaAs mixture. The growth temperature interval was (803–802) ◦C with 0.5 ◦C/min cooling speed. The epitaxial layers were not doped intentionally. AlAs mole fraction x in the AlxGa1-xAs semiconductor was chosen as x = 0; 0.15; and 0.3. Optical microscope was used to measure the thickness of the grown layers by taking a cross-sectional view of the structure fraction. Electrical characteristics of the layers were measured using Van der Pauw technique at room temperature. The main parameters of the layers are presented in Table1. Sensors 2021, 21, x FOR PEER REVIEW 3 of 17
SensorsSensors 20212021,, 2121,, x 4487 FOR PEER REVIEW 3 3of of 17 17
Table 1. Parameters of the epitaxially grown AlxGa1-xAs layers.
Table 1. Parameters of the epitaxially grown AlxGa1-xAs layers. Table 1. Parameters ofLayer the epitaxially Electron grown Al xGa1-xElectronAs layers. Electrical Sheet AlAs Mole ThicknessLayer DensityElectron n, MobilityElectron μ, ElectricalResistivity ResistanceSheet AlAsFraction Mole x Layer Electron Electron Electrical Sheet AlAs Mole Thicknesst, μm Densitycm−3 n, Mobilitycm2/(V·s) μ, Resistivityρ, Ω·cm ResistanceRsh, Ω/ Fraction x Thickness t, Density n, Mobility µ, Resistivity ρ, Resistance Fraction x ×−33 16 22 0 t, µμm0.8m 5.5cm 10 cmcm1820/(V·s)/(V·s) ρ,Ω Ω·cm·cm6 RRshsh, 780,Ω/Ω/ × 1616 0.1500 0.8 0.80.8 5.55.52.0 × 10 101016 182018201680 6 196 780 2330780 0.150.150.3 0.8 0.80.5 2.02.06.0 × × 10 10101615 168016801500 19 19 70 2330 13880 2330 0.30.3 0.5 0.5 6.06.0 × 101015 15001500 70 70 13880 13,880 Schematic view of the planar microwave diode is presented in Figure 1a, and its cross-sectionalSchematicSchematic viewview of ofis theshown the planar planar in microwaveFigure microwave 1b. Simple diode diode is fabrication presented is presented inprocedure Figure in Figure1a, of and the 1a, its diode and cross- its is crosssectionalsketched-sectional viewin Figure isview shown 1c–e.is shown in Before Figure in the1Figureb. Simplesample 1b. fabricationSimpleprocessing, fabrication procedure the grown procedure of wafers the diode of were the is sketcheddioderinsed isin sketchedinacetone Figure 1inandc–e. Figure methanol. Before 1c– thee. Before sampleFirst, photolithograpthe processing, sample processing, thehy grown was performedwafersthe grown were usingwafers rinsed positive were in acetone rinsed Shipley and in acetonemethanol.S1805 photoresist and First, methanol. photolithography to form First, rectangular photolithography was performed105 × 60 μ wasm using2 AlGaAs performed positive mesas Shipley using (see S1805 positiveFigure photoresist 1c). Shipley Phos- S1805tophorus-based form photoresist rectangular etchant to 105 form of× special 60rectangularµm2 composition,AlGaAs 105 mesas× 60 H μm (see3O24 :HAlGaAs Figure2O2:H12c). Omesas = Phosphorus-based 1:4:45 (see [29], Figure was 1 chosenc). etchant Phos- on pofpurpose,horus special-based creating composition, etchant mesa of Hspecialwith3O4 :Hfl composition,at2O slopes;2:H2O =this 1:4:45 H is3O important4:H [292O],2 was:H2O to chosen = protect1:4:45 on [29], the purpose, wasmetal chosen creatingterminal on purpose,mesafrom withcracking creating flat slopes;on themesa mesa’s this with is importantslope.flat slopes; The toetched this protect is mesa’s important the metal slope toterminal was protect controlled from the crackingmetal using terminal “Dektak on the frommesa’s6M” cracking(Veeco slope. Metrology Theon the etched mesa’s LLC, mesa’s slope. Plai slope nview,The was etched NY, controlled mesa’sUSA) stylus usingslope “Dektakwasprof ilometer.controlled 6M” (Veeco The using angle Metrology “Dektak of the 6M”LLC,slope (Veeco Plainview,ranged Metrology between NY, USA) 21°LLC, and stylus Plainview, 24° profilometer.independently NY, USA) The of stylus orientation angle profilometer. of the of slopethe mesas, The ranged angle with between respectof the slope21to◦ crystallographicand ranged 24◦ independentlybetween directions 21° and of 24° of orientation theindependently semico ofnductor the of mesas, orientation substrate. with respect ofSecond, the mesas, to photolithography crystallographic with respect todirectionswas crystallographic carried of out the using semiconductor direction images ofreversal the substrate. semiconductor MicroChemicals Second, substrate. photolithography AZ 5214E Second, photoresist photolithography was carriedto open out the wasusingwindows carried image for out reversal the using deposition MicroChemicalsimage reversalof metallic MicroChemicals AZ contacts. 5214E photoresistThe width AZ 5214E tod of open thephotoresist thetip of windows the to left open forcontact the the windowsdeposition(see Figure for of1a) the metallic varied deposition in contacts. the of range metallic The of width1–4 contacts. μmd inof different theThe tipwidth of MW the d of leftdiodes. the contact tip Metallic of the (see left Figurecontacts contact1a) of (seevariedthe Figurediodes in the 1a )range werevaried of fabricatedin 1–4 the µrangem in byof different 1–means4 μm MW in ofdifferent diodes.thermal MW Metallic successivediodes. contacts Metallic evaporation of contacts the diodes ofof thewereNi:Au:Ge:Ni:Au diodes fabricated were by metals means fabricated of ofrespective thermal by meansthickne successive ss of 5:200:100:75:100 evaporation thermal successive of nm Ni:Au:Ge:Ni:Au onto evaporationa photo-resistive metals of Ni:Au:Ge:Ni:Auofmask respective using “VAKSIS thickness metals PVD of 5:200:100:75:100 respective Vapor-5S_Th” thickness nm (Vak onto 5:200:100:75:100sis, a photo-resistiveAnkara, Turkey) nm onto mask (Figure a usingphoto 1d).- “VAKSISresistive Compo- maskPVDsition Vapor-5S_Th”using of Ge–Ni–Au-based “VAKSIS (Vaksis, PVD Vapormeta Ankara,ls- 5S_Th”was Turkey) chosen (Vaksis, (Figure because Ankara,1d). it Composition assures Turkey) good (Figure of Ohmic Ge–Ni–Au-based 1d ).contacts Compo- to sitionmetalsn-GaAs of was Geand– chosenNi AlGaAs–Au because-based [30]. metals itThe assures contact was good chosen patterns Ohmic because were contacts formedit assures to n-GaAs using good and the Ohmic AlGaAslift-off contacts technique [30]. The to ncontactwith-GaAs consequent patternsand AlGaAs were annealing [30]. formed The of using thecontact structures the patterns lift-off at technique 430were °C formed in withhydrogen consequentusing atmosphere the lift annealing-off techniquein the of tube the ◦ withstructuresfurnace consequent (Figure at 430 1e).annealingC inAt hydrogenthis ofstage, the atmosphere structuresit is worth atnoting, in 430 the °C tubethe in fabricationhydrogen furnace (Figure atmosphere procedure1e). At of in this the the stage,planar tube furnaceitMW is worth diodes (Figure noting, situated 1e). the At onfabricationthis the stage, semi-insulating it procedure is worth noting, of substrate the planar the fabrication was MW completed, diodes procedure situated and ofsuch on the the samples planar semi- MWinsulatingwere diodes handy substrate situated for the wasonmeasurements the completed, semi-insulating on and a suchprobe substrate samples station. was were In completed,contrast, handy for the and the dual measurementssuch MW samples diodes wereonthat a probewerehandy based station.for t heon measurements Inlow-resistivity contrast, the onGaAs dual a probe MWsubstrate diodesstation. [27] that In additionally contrast, were based the needed ondual low-resistivity MW to be diodes trans- thatGaAsferred were substrate onto based a dielectric [ 27on] additionallylow -polyimideresistivity needed GaAsfilm toin substrate beorder transferred to [27]avoid additionally onto electrical a dielectric shunting needed polyimide tothrough be filmtrans- the in ferredorderlow-resistivity toonto avoid a dielectric electrical GaAs substrate. polyimide shunting Micrographic through film in theorder low-resistivity view to avoid of the electrical finished GaAs substrate. shunting planar MWMicrographic through diode the on lowviewsemi-insulating-resistiv of theity finished GaAs substrate planar substrate. is MW presented Micrographic diode on insemi-insulating Figure view 2. of the substrate finished is planar presented MW in diodeFigure on2 . semi-insulating substrate is presented in Figure 2.
FigureFigure 1.1. SchematicSchematic toptop ((aa)) andand cross-sectionalcross-sectional ( b(b)) views views of of the the planar planar microwave microwave diode. diode. Successive Succes- sive fabrication steps of the diode are shown in parts (c–e). Figurefabrication 1. Schematic steps of thetop diode(a) and are cross shown-sectional in parts (b) ( cviews–e). of the planar microwave diode. Succes- sive fabrication steps of the diode are shown in parts (c–e).
Sensors 2021, 21, x FOR PEER REVIEW 4 of 17 Sensors 2021, 21, 4487 4 of 17
Figure 2. Micrograph of the planar microwave diode on semi-insulating GaAs. Figure 2. Micrograph of the planar microwave diode on semi-insulating GaAs. Probe stations were used to measure both dc and high-frequency parameters of the microwaveProbe stations diodes. were The used choice to measure of these instrumentsboth dc and washigh-frequency advantageous parameters in collecting of the more microwaveinformative diodes. statistical The choice data of of the these diodes. instru Thements current–voltage was advantageous (IV) characteristics in collecting were more mea- informativesured using statistical Süss Micro data Tec of probethe diodes. station The EP6 current–voltage (FormFactor, Inc., (IV) Livermore, characteristics CA, USA) were and measuredAgilent using E5270B Süss precision Micro measurementTec probe station equipment EP6 (FormFactor, (Agilent Technologies, Inc., Livermore, Inc., CA, Santa USA) Clara, andCA, Agilent USA). E5270B Voltage–power precision (VP) measurement characteristics equi ofpment the diodes (Agilent in the Technologies, Ka frequency Inc., range Santa were Clara,measured CA, USA). using Voltage–power Cascade Microtech (VP) (FormFactor, characteristics Inc., of Livermore, the diodes CA, in USA) the K high-frequencya frequency rangeprobe were station, measured and probesusing Cascade ACP40-A-GS-250 Microtech were (FormFactor, used to connectInc., Livermore, the investigated CA, USA) MW high-frequencydiode to the measurement probe station, device. and probes SHF BT45 ACP40-A-GS-250 broadband bias were tee used separated to connect the detected the in- dc vestigatedvoltage signalMW diode from to the the microwave measurement signal. device. Application SHF BT45 of the broadband probe stations bias tee gave separated additional theoptions detected in dc the voltage investigation signal from of electrical the microwave properties signal. of the Application planar microwave of the probe diodes, sta- i.e., tionswe gave could additional measure options the characteristics in the investigation depending of electrical on light illuminationproperties of andthe planar temperature. mi- crowaveMost measurements diodes, i.e., wewere could carried measure out the at characteristics room temperature, depending but use on of light commercially illumination avail- andable temperature. Peltier modulus Most measurements let us perform were both ca dcrried and out high-frequency at room temperature, measurements but use up of to ◦ commercially80 C temperature. available Peltier modulus let us perform both dc and high-frequency meas- urementsThe up sequence to 80 °C oftemperature. the study of the fabricated planar MW diodes was as follows. First, separateThe sequence MW diodes of the were study tested of the in fabricated their array planar on the MW semiconductor diodes was substrate.as follows. Electrical First, separateresistance MW at diodes zero applied were tested voltage, in their which arra wasy on evaluated the semiconductor from the measured substrate. IV Electrical characteris- resistancetics. Then at zero the voltage applied responsivity voltage, which of thewas diodes evaluated was from measured the measured at f = 30 GHzIV character- frequency. istics.The Then level the of the voltage microwave responsivity power wasof the chosen diodes to was be in measured the range at where f = 30 the GHz detected frequency. voltage Thedepended level of the linearly microwave on the power.power was Median chosen values to be of thein the voltage range responsivity where the detected and electrical volt- re- agesistance depended of the linearly diodes on were the calculatedpower. Median after statisticalvalues of processingthe voltage of responsivity the measured and data. elec- For tricalfurther, resistance more detailedof the diodes investigation, were calculated the diodes after having statistical electrical processing parameters of the close measured to the me- data.dian For values further, were more selected detailed from investigation, all three groups the of diodes different having AlAs moleelectrical fraction-containing parameters x closesamples to the (havingmedian values= 0, 0.15 were and selected 0.3). Tofrom find all out three the groups influence of different of illumination AlAs mole on the electrical parameters of the MW diodes, the IV and VP (at f = 30 GHz) characteristics were fraction-containing samples (having x = 0, 0.15 and 0.3). To find out the influence of illu- measured in the dark and under photo lamp Eiko EKE21V150W (color temperature 3240 K) mination on the electrical parameters of the MW diodes, the IV and VP (at f = 30 GHz) light illumination. Frequency dependence of the voltage responsivity was measured within characteristics were measured in the dark and under photo lamp Eiko EKE21V150W the K frequency range. The responsivity of the diodes in wider frequency range was (color temperaturea 3240 K) light illumination. Frequency dependence of the voltage re- performed by fitting the experimental responsivity data and using the phenomenological sponsivity was measured within the Ka frequency range. The responsivity of the diodes in equation of the dependence of voltage responsivity on non-linearity parameters of IV wider frequency range was performed by fitting the experimental responsivity data and characteristics of the diodes. The hardiness of the diodes to withstand maximum power of using the phenomenological equation of the dependence of voltage responsivity on non- the MW radiation was evaluated using the value of the highest measured applied voltage linearity parameters of IV characteristics of the diodes. The hardiness of the diodes to at which the IV characteristic still did not begin changing.
Sensors 2021, 21, 4487 5 of 17
3. Results and Discussion Three batches of different kinds of the diodes, DD0, DD15 and DD30, were named according to the percentage of AlAs mole fraction in the AlxGa1-xAs semiconductor com- pound. The fabricated diodes also had different widths d of the small area contact, as follows: 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0 micrometers. Each batch contained 480 diodes. The output of the operative diodes in the DD0, DD15 and DD30 batches was 83%, 78% and 46%, respectively. The two-fold decrease in the number of the operative diodes in the case of DD30 can be explained by a significant increase in the diode’s resistance when the molar fraction of AlAs exceeds 20%, and when the deep recombination centers (DX centers) in the AlGaAs compound begin to appear [31]. A negative potential of the detected voltage was predominantly measured on the small area contact (left contact in Figures1 and2 ). Let us label this polarity detected voltage as a Schottky voltage. However, the voltage of opposite polarity was detected across some operative DD0 and DD15 diodes, and it amounted to 7% and 4%, respectively. Let us call this detected voltage a TEMF voltage, because its polarity is inherent to the thermoelectromotive force of the hot carriers measured across the mi- crowave diodes having perfect Ohmic contacts [19]. The box charts of the statistical results are presented in Figure A1 of AppendixA. The median value of the voltage responsivity of the diodes can be approximated by its linear dependence on the width of the contact d, as follows: RV = RV (0) − b·d, (1)
where RV(0) stands for the extrapolated value of the voltage responsivity at d = 0, and b is the slope of the approximation line. The voltage responsivity of the diodes DD0 is almost independent on the contact width, while the linear dependence of the voltage responsivity of the planar diodes on the base of AlGaAs is stronger, and the slope b increases with the increase in the AlAs mole fraction x in the semiconductor compound. It should be noted that the measurements of the detected voltage were performed using high impedance loading. If a 50 Ohms load was used, then the voltage responsivity would decrease. In the case of the DD15 diode, the responsivity decreases from 100 V/W down to 3 V/W. However, in the case of the 50 Ohms load, the DD15 diode can be used to detect pulsed nanosecond- long microwave signals. The electrical resistance R0 at zero voltage is scattered around its median value in the case of the DD0 and DD15 diodes, and it follows linear dependence close to Equation (1), with coefficient bR for the DD30 diodes. Table2 summarizes the values of the voltage responsivity and electrical resistance of the planar microwave diodes in respect of the width of their narrow contact.
Table 2. Parameters of the linear dependence of voltage sensitivity and electrical resistance on diodes’ contact width.
Voltage Electrical Slope b, Slope b , Diode Responsivity at Resistance at R V/(W·µm) kΩ/µm d = 0 RV(0), V/W d = 0 R0(0), kΩ DD0 0.9 ± 0.16 0.02 ± 0.06 1.0 ± 0.1 0.03 ± 0.04 DD15 98 ± 10 20.5 ± 3.9 2.16 ± 0.33 0.2 ± 0.12 DD30 207 ± 19 40 ± 7 15.6 ± 3.1 2.6 ± 1.2
We tried to approximate the statistical data of our experiments using the following various distribution functions: normal, lognormal, and Weibull. The last function was the best to fit the experimental results of the voltage responsivity and electrical resistance. Figure A2 in AppendixA depicts histograms of the experimental data of the voltage responsivity and electrical resistance of all three batches of the planar microwave diodes, DD0, DD15, and DD30. The solid lines in Figure A2 denote the Weibull distribution function. Its parameters, the shape β and the scale η, are presented in the legends of Figure A2 in AppendixA, and are summarized in Table3. Sensors 2021, 21, 4487 6 of 17
Table 3. Shape and scale parameters of the Weibull distribution function for experimental voltage
responsivity RV and electrical resistance R0 data.
Shape β Scale η Diode for RV for R0 for RV for R0 DD0 1.0 1.47 1.72 1.34 DD15 0.77 0.89 76.8 1.28 DD30 1.51 0.89 151 19.2
A comparison of the data presented in Tables2 and3 shows a qualitative correlation between the voltage responsivity and electrical resistance median data and scale parameter η, which is derived from the data approximation by the Weibull distribution function (Equation (A1) in AppendixA). For a more detailed, further investigation, representative diodes having median values of the voltage responsivity RV and electrical resistance R0 were selected from all three batches. Electrical parameters of the selected planar microwave diodes measured under white-light illumination and in the dark are presented in Table4.
Table 4. Voltage responsivity RV and electrical resistance R0 of the representative illuminated and darkened planar microwave diodes.
Illuminated Dark Diode RV, V/W R0, kΩ RV, V/W R0, kΩ DD0 1.2 1.1 1.7 1.8 DD15 100 1.70 105 1.8 DD30 160 14.3 150 14.9
The IV characteristics of the white-light-illuminated diodes are presented in Figure3. The forward current flows through the diodes when a positive potential of the voltage is applied to the small area contact, as is shown in Figure3. The IV characteristic of the DD0 diodes is quasilinear, while it becomes nonlinear in the case of the DD15 and DD30 diodes. The nonlinearity is more strongly pronounced for the DD30 diode, which is fabricated on the base of higher resistivity Al0.3Ga0.7As semiconductor material. The asymmetry of the current–voltage characteristics also increases with the mole fraction x. Correlation of the IV characteristic’s asymmetry with the diode’s voltage responsivity allows us to expect an appropriate increase in responsivity as the diode demonstrates a more asymmetric IV characteristic. A reliable relation exists between the IV characteristic’s asymmetry and the diode’s voltage responsivity. When the relaxation of the average energy of hot electrons in a semiconductor can be neglected, and the conductivity current exceeds the displacement current, then the voltage responsivity of the microwave diode with n-n+ junction can be expressed as [20]. Rr − R f R = , (2) V 2U
where Rr and Rf denote the electrical resistance in the reverse and forward, respectively, direction of the applied bias voltage U. Sensors 2021, 21, x FOR PEER REVIEW 7 of 17
SensorsSensors 20212021, ,2121, ,x 4487 FOR PEER REVIEW 7 7of of 17 17
Figure 3. Dependence of electrical resistance of the illuminated planar microwave diodes on ap- plied bias voltage. IV characteristics of the diodes are presented in the inset.
𝑅 −𝑅 𝑅 = , (2) 2𝑈 Figure 3. Dependence of electrical resistance of the illuminated planar microwave diodes on applied Figurewhere 3. R Dependencer and Rf denote of electrical the electrical resistance resistance of the illuminated in the reverse planar andmicr owaveforward, diodes respectively, on ap- bias voltage. IV characteristics of the diodes are presented in the inset. plieddirection bias voltage. of the appliedIV characteristics bias voltage of the U diodes. are presented in the inset. The dependence of the detected voltage on microwave power, i.e.,i.e., VPVP characteristics of the planar microwave diodes, was was measured measured𝑅 −𝑅 using using a a high-frequency high-frequency probe probe station station at at f 𝑅 = , (2) f= =30 30 GHz GHz frequency. frequency. The The characteristics characteristics are are presented presented2𝑈 in in Figure Figure 4.4.
where Rr and Rf denote the electrical resistance in the reverse and forward, respectively, direction of the applied bias voltage U. The dependence of the detected voltage on microwave power, i.e., VP characteristics of the planar microwave diodes, was measured using a high-frequency probe station at f = 30 GHz frequency. The characteristics are presented in Figure 4.
Figure 4. Voltage–power characteristics of the planar microwave diodes measured under white-light Figure 4. Voltage–power characteristics of the planar microwave diodes measured under white- illumination at f = 30 GHz frequency. Lines are guides for the eye of linear dependence. light illumination at f = 30 GHz frequency. Lines are guides for the eye of linear dependence. The microwave power on the graph is the incident power (not the absorbed one), i.e., the power supplied to the HF probe. The voltage standing wave ratio (VSWR) was not measured in the case of the experiments carried out using the probe station. It can be conducted when the planar MW diode is mounted into a waveguide. At low MW power, Figure 4. Voltage–power characteristics of the planar microwave diodes measured under white- lightthe detectedillumination voltage at f = 30 of GHz all the frequency. microwave Lines diodes are guides follows for the the eye linear of linear law. dependence. However, the VP characteristics become different at higher power values; they turn sublinear in the case of
Sensors 2021, 21, x FOR PEER REVIEW 8 of 17
The microwave power on the graph is the incident power (not the absorbed one), i.e., the power supplied to the HF probe. The voltage standing wave ratio (VSWR) was not measured in the case of the experiments carried out using the probe station. It can be con- Sensors 2021, 21, 4487 ducted when the planar MW diode is mounted into a waveguide. At low MW power,8 the of 17 detected voltage of all the microwave diodes follows the linear law. However, the VP characteristics become different at higher power values; they turn sublinear in the case of thethe DD15 DD15 and and DD30 DD30 diodes, diodes, while while the the DD0 DD0 diode’s diode’s characteristic characteristic remains remains linear, linear, within within thethe measured measured MW MW power power range. range. The The difference difference in in the the VP VP characteristics characteristics can can be be partly partly explainedexplained by by the the difference difference in in the the asymmetry asymmetry of of the the IV IV characteristics. characteristics. Figure Figure 55 depictsdepicts the the asymmetryasymmetry versus versus applied applied voltage voltage calculated calculated using using Equation Equation (2). (2).
FigureFigure 5. 5. DependenceDependence of of the the IV IV asymmetry asymmetry of of the the planar planar microwave microwave diodes diodes on on applied applied bias bias volt- voltage. age. The asymmetry of the DD0 diode’s IV characteristic is the lowest, and that of the DD15The diode asymmetry is higher of bythe almost DD0 diode’s two orders IV characteristic of magnitude. is Thisthe lowest, correlates and well that with of the the DD15VP experimental diode is higher results by (Figurealmost4 two), which orders found of magnitude. that the voltage This responsivitycorrelates well of thewith DD15 the diode is higher, by two orders of magnitude, as compared to that of the DD0 diode. On the VP experimental results (Figure 4), which found that the voltage responsivity of the DD15 other hand, the IV asymmetry of the DD30 diode is higher, by two orders of magnitude, diode is higher, by two orders of magnitude, as compared to that of the DD0 diode. On than that of the DD15 diode, while the voltage responsivity is higher by only two times. the other hand, the IV asymmetry of the DD30 diode is higher, by two orders of magni- Poor coincidence between the IV asymmetry and voltage responsivity data of the DD15 tude, than that of the DD15 diode, while the voltage responsivity is higher by only two and DD30 diodes can be explained by weaker absorption of the incident MW radiation times. Poor coincidence between the IV asymmetry and voltage responsivity data of the by the DD30 diode, due to the higher value of its electrical resistance (see Table4). The DD15 and DD30 diodes can be explained by weaker absorption of the incident MW radi- deviation from the linear law (VP characteristics of diodes DD15 and DD30) most probably ation by the DD30 diode, due to the higher value of its electrical resistance (see Table 4). results from the decrease in the diodes’ IV asymmetry at a lower applied voltage than in The deviation from the linear law (VP characteristics of diodes DD15 and DD30) most the case of the DD0 diode. probably results from the decrease in the diodes’ IV asymmetry at a lower applied voltage As the data presented in Table4 show, white-light illumination has a different impact than in the case of the DD0 diode. on the electrical parameters of the microwave diodes. The DD0 diodes are more sensitive to theAs illumination;the data presented both theirin Table voltage 4 show, responsivity white-light and illumination electrical ha resistances a different decrease impact by onabout the electrical 30% when parameters the light is of put the on. microwave The change diodes. in the The parameters DD0 diodes of the are DD15 more andsensitive DD30 todiodes the illumination; is lower, by aboutboth their one order voltage ofmagnitude, responsivity as and compared electrical to the resistance DD0 diode. decrease A weaker by about 30% when the light is put on. The change in the parameters of the DD15 and DD30 dependence of R0 on illumination is associated with the formation of different densities diodesof activation is lower, energy by about levels one at theorder metal–semiconductor of magnitude, as compared interface, withto the an DD0 increase diode. in theA weakerAlAs moledependence fraction ofx [R320 ]on and illumination also with ais different associated nature with of the deep formation traps in of the different bulk of densities of activation energy levels at the metal–semiconductor interface, with an in- the AlxGa1-xAs ternary semiconductor depending on x [33]. It is worth noting that the creasevoltage in the responsivity AlAs mole of fraction the diode x [32] DD30 and increasesalso with witha different illumination. nature of The deep IV traps asymmetry in the of the planar microwave diodes also correlates with the voltage responsivity, considering the impact of illumination; the light-induced change in the asymmetry decreases with the rise in x in the AlxGa1-xAs compound. This correlation also supports the presumption that the detected voltage arises due to the rectification of microwave currents across the quasi–ohmic metal–semiconductor junctions of the diodes. Sensors 2021, 21, 4487 9 of 17
Actually, the developed planar microwave diodes are a composition of two quasi- ohmic metal–semiconductor junctions. Therefore, a natural expectation is to describe the performance of the diodes using equations that are inherent to a metal–semiconductor junction. We fitted the experimental data of the IV characteristics of the diodes by the formula of a Schottky junction [34], as follows:
eU I = I exp − 1 , (3) s nkT
where I and Is are the current and the saturation current, respectively, U stands for the applied bias voltage, n marks the non-ideality factor, k notes the Boltzmann constant, and T is the diode temperature. The saturation current Is and the non-ideality factor n were chosen as the fitting parameters. The fitting was performed using IV data not exceeding 0.15 V of the applied voltage. The fitting parameters of the diodes are presented in Table5.
Table 5. Fitting parameters of the IV characteristics approximated with Equation (3).
Saturation Current Non-Ideality Factor Diode Is,A n DD0 (3.5 ± 2.1) × 10−2 1520 ± 15 DD15 (2.9 ± 0.12) × 10−4 20 ± 0.7 DD30 (2.3 ± 0.24) × 10−5 11.2 ± 0.9
The experimental IV characteristics of the diodes DD15 and DD30 only can be satisfac- torily fitted using Equation (3). The best fitting results are achieved for the DD15 diodes, when the fitting parameters are scattered by ~4%, while the parameters of the diodes DD30 are scattered by ~10%. No satisfactory agreement between the IV experimental data and the approximation by Equation (3) was obtained for the DD0 diodes. An unusually high value of the non-ideality factor can be explained by the metal-semiconductor junction being far from the ideal Schottky junction. Ge/Ni/Au metals annealed at high tempera- tures make the quasi–ohmic metal-semiconductor junction have such a high value for the non-ideality factor. If the IV characteristic of a microwave diode is described by Equation (3), then its voltage responsivity can be expressed as [35].
γRjkabs RV = h i , (4) Rs Rs 2 2 1 + 1 + + ωC RsR Rj Rj j j
= d2 I dI where γ dU2 / dU denotes the nonlinearity of the IV characteristic, Rj and Cj stand for the barrier resistance and capacitance of a metal–semiconductor junction, respectively, Rs is the series resistance of the diode, ω is the angular frequency of a microwave signal, and kabs notes the part of microwave radiation absorbed by the diode. Some parameters in the formula can be derived from the IV characteristic of the diode, namely, the non-linearity parameter γ and the electrical resistances of the diode. The barrier resistance Rj can be cal- culated by means of subtracting the series resistance Rs from the experimentally measured resistance R0, at zero applied voltage. However, in the case when the IV characteristic of the diode is far from the characteristic of an ideal Schottky junction (the case of the DD0 diode), the experimental estimation of Rs becomes complicated. Therefore, we calculated the geometrical series resistance of the diode using the experimental value of the sheet resistance Rsh of the epitaxial layer (see Table1), with the following formula
(D − d)ln D = d Rs Rsh D . (5) 4Lln d + π(D − d) Sensors 2021, 21, x FOR PEER REVIEW 10 of 17
diode), the experimental estimation of Rs becomes complicated. Therefore, we calculated the geometrical series resistance of the diode using the experimental value of the sheet resistance Rsh of the epitaxial layer (see Table 1), with the following formula 𝐷 (𝐷−𝑑)𝑙𝑛 𝑅 =𝑅 𝑑 . Sensors 2021, 21, 4487 𝐷 10(5) of 17 4𝐿𝑙𝑛 +𝜋(𝐷−𝑑) 𝑑 Here, we used geometrical parameters of the diode d, D and L, which are shown in the insetHere, of Figure we used 1. The geometrical series resistance parameters of the of thediodes diode DD0,d, D DD15and L and, which DD30, are calculated shown in the usinginset Equation of Figure (5),1 .was The 190 series Ω, 580 resistance Ω, and of3460 the Ω diodes, respectively, DD0, DD15 taking and the DD30, small calculatedcontact widthusing d = Equation1 μm, contact (5), was length 190 LΩ =, 58010 μΩm,, andand 3460gap widthΩ, respectively, D = 15 μm. taking Such theevaluation small contact of the widthseries dresistance= 1 µm, contactof the diodes length wasL = 10 onlyµm, successful and gap widthin the Dcase= 15 ofµ them. SuchDD30 evaluation diode. In of the thecase series of the resistance DD15 and of theDD0 diodes diodes, was the only calculated successful value in the was, case respectively, of the DD30 1.5 diode. and In 5 the timescase lower of the than DD15 the andexperimental DD0 diodes, one. the The calculated mismatch value between was, the respectively, calculated 1.5 and and exper- 5 times imentallower values than thecan experimentalbe associated one.with Thea stronger mismatch influence between of the the surface calculated states, and as experimental the AlAs molevalues fraction can bein associatedthe semiconductor with a stronger compound influence is reduced. of the surface The absorbed states, as themicrowave AlAs mole powerfraction (absorption in the semiconductorcoefficient kabs), as compound well as the is barrier reduced. capacitance The absorbed Cj of the microwave diode can power be found(absorption from Equation coefficient (4), andkabs), can as wellbe used as the to barrierapproximate capacitance the experimentalCj of the diode dependence can be found of thefrom voltage Equation responsivity (4), and canof the be planar used to micr approximateowave diodes the experimentalon frequency, dependence as is shown ofin the Figurevoltage 6. responsivity of the planar microwave diodes on frequency, as is shown in Figure6.
. FigureFigure 6. Experimental 6. Experimental frequency frequency dependence dependence of volt ofage voltage responsivity responsivity of the planar of the microwave planar microwave diodesdiodes (dots) (dots) in the in theKa frequency Ka frequency range. range. Solid Solid lines lines mark mark the thedependences dependences in wider in wider frequency frequency range rangecalculated calculated using using Equation Equation (4). (4).
TheThe parameters parameters used used to fit tothe fit calculated the calculated RV dataRV todata the experimental to the experimental points are points pre- are sentedpresented in Table in 6. Table 6.
Table 6. Experimental and approximation parameters of the planar microwave diodes.
R (1 GHz), R (1 THz), γ,V−1 R , kΩ R , kΩ C , fF k ,% V V j s j abs V/W V/W Diode Approximation Approximated Experimental Parameters Parameters Parameters DD0 0.07 0.9 0.19 12 10 1.8 0.003 DD15 2.2 1.23 0.54 10 44 290 0.16 DD30 8.5 10.9 3.6 1 1.4 370 0.35
Such an approximation technique opens the possibility of evaluating the high-frequency capacitance of the microwave diodes (as small as in the femtofarad range), as well as finding the microwave power absorbed in the diode. Moreover, the performance of the microwave diodes in the THz frequency range can also be forecasted. As Table6 shows, the values of the voltage responsivity of the DD15 and DD30 diodes let us expect their application, even Sensors 2021, 21, x FOR PEER REVIEW 11 of 17
Table 6. Experimental and approximation parameters of the planar microwave diodes.
RV (1 GHz), RV (1 THz), γ, V−1 Rj, kΩ Rs, kΩ Cj, fF kabs,% V/W V/W Diode Approximation Parame- Approximated Experimental Parameters ters Parameters DD0 0.07 0.9 0.19 12 10 1.8 0.003 DD15 2.2 1.23 0.54 10 44 290 0.16 DD30 8.5 10.9 3.6 1 1.4 370 0.35
Such an approximation technique opens the possibility of evaluating the high-fre- quency capacitance of the microwave diodes (as small as in the femtofarad range), as well as finding the microwave power absorbed in the diode. Moreover, the performance of the microwave diodes in the THz frequency range can also be forecasted. As Table 6 shows, the values of the voltage responsivity of the DD15 and DD30 diodes let us expect their Sensors 2021, 21, 4487 11 of 17 application, even in the terahertz frequency range. Of course, our planar diodes are weak competitors with the Schottky diodes, in the sense of responsivity. However, the diodes havingin the terahertz responsivity frequency range. of the Of course,order our of planar (0.1–1.0) diodes areV/W weak can competitors be used with to detect electromagnetic the Schottky diodes, in the sense of responsivity. However, the diodes having responsivity radiationof the order in of THz (0.1–1.0) frequency V/W can be range used to [21,23]. detect electromagnetic radiation in THz frequencyThe range relative [21,23 ].change in voltage responsivity with temperature of all three types of diodesThe is relative depicted change in in voltageFigure responsivity 7. with temperature of all three types of diodes is depicted in Figure7.
FigureFigure 7. 7.Temperature Temperature dependence dependence of relative voltage of relative responsivity voltage change inresponsivity the planar microwave change in the planar micro- diodes. wave diodes. Since the planar microwave diode is composed of two metal–semiconductor junc- tions, connected in series in opposite directions (see Figure1), this way, the resulting IV characteristicSince ofthe such planar a planar microwave diode should be diode made of is backward comp branchesosed of of two both junc-metal–semiconductor junc- tions,tions. Therefore,connected by following in series the equationin opposite below of directio the saturationns (see current Figure of a Schottky 1), this way, the resulting IV junction [34]: characteristic of such a planar diode2 − e ψshouldms be made of backward branches of both junc- I = A T e kT (6) tions. Therefore, by followings theR equation below of the saturation current of a Schottky (here AR is the Richardson constant, ψms marks the barrier height of a Schottky junction), junctionone should [34]: expect a decrease in electrical resistance of the planar microwave diode with higher temperatures. The temperature dependence of the electrical resistance of the diodes at zero applied voltage is presented in Figure8, in solid points, and the values of the (6) 𝐼 = 𝐴 𝑇 𝑒 non-linearity coefficient γ of the IV characteristics are also depicted here, in open dots. (here AR is the Richardson constant, ψms marks the barrier height of a Schottky junction), one should expect a decrease in electrical resistance of the planar microwave diode with
Sensors 2021, 21, x FOR PEER REVIEW 12 of 17
higher temperatures. The temperature dependence of the electrical resistance of the di- Sensors 2021, 21, 4487 odes at zero applied voltage is presented in Figure 8, in solid points, and the values of12 the of 17 non-linearity coefficient γ of the IV characteristics are also depicted here, in open dots.
FigureFigure 8. Temperature 8. Temperature dependence dependence of electrical of electrical resistan resistancece of the of theplanar planar microwave microwave diodes diodes at zero at zero appliedapplied voltage voltage (solid (solid dots), dots), and and temperature temperature dependence dependence of ofthe the non-linearity non-linearity coefficient coefficient γ γofof the the IV IV characteristics of the diodes at zero voltage (open dots). The The colors colors correspond correspond to to the the same same type type of of the diodes as in Figure 7. the diodes as in Figure7.
TheThe electrical electrical resistance resistance of ofthe the diodes diodes DD15 DD15 and and DD30 DD30 drops drops down down as as their their temper- tempera- atureture rises. rises. This This characteristic characteristic agrees agrees with with saturation saturation current current dependence dependence on ontemperature, temperature, as asdescribed described by byEquation Equation (6). (6).However, However, the dependence the dependence of the of DD0 the DD0diode’s diode’s resistance resistance on temperatureon temperature is more is morestipulated stipulated by the by bulk the bulk resistance resistance of GaAs, of GaAs, and and therefore therefore the the DD0 DD0 diode’sdiode’s resistance resistance increases increases with with temperature. temperature. Moreover, Moreover, the sign the of sign the of nonlinearity the nonlinearity co- efficientcoefficient of its ofIV its characteristic IV characteristic is also is opposite also opposite to that to of that the ofdiodes the diodes DD15 and DD15 DD30 and (see DD30 Table(see 6).Table As 6Figures). As Figures7 and 87 show, and8 ashow, good acorrelation good correlation exists between exists between all three allof threethe pa- of rameters,the parameters, i.e., voltage i.e., responsivity, voltage responsivity, resistance resistance,, and nonlinearity and nonlinearity coefficient, coefficient, in their in de- their pendencedependence on temperature. on temperature. The detection The detection properties properties of the diodes of the on diodes the base on of the the base AlxGa of1- the xAsAl ternaryxGa1-xAs semiconductor ternary semiconductor (diodes DD15 (diodes and DD15DD30) and are DD30)determined are determined by the rectification by the rec- of tificationmicrowave of currents microwave in the currents integrate in the planar integrate dual microwave planar dual diode, microwave which diode,is composed which is of composedtwo metal–semiconductor of two metal–semiconductor quasi–ohmic quasi–ohmicjunctions that junctions are connected that are in connected series in oppo- in series sitein directions. opposite directions. Finally,Finally, we we carried carried out out an an evaluation evaluation of of another another important important characteristic characteristic of of a a sensor. sensor. It It isis the the ability ability of of the the microwave microwave diode diode to to withstand withstand the the impact impact of ofhigh-power high-power microwave microwave radiation.radiation. In the In thefirst first step, step, we measured we measured the maximum the maximum forward forward bias voltage bias voltage (positive (positive po- tentialpotential is applied is applied to the to small the small area areacontact), contact), at which at which the thenon-reversible non-reversible changes changes in the in the IV IV characteristiccharacteristic still still do donot not take take place. place. Then, Then, the the maximum maximum electrical electrical power power absorbed absorbed by bythe the diodediode is calculated is calculated and and the the maximum maximum incident incident MW MW power power is found, is found, by bytaking taking the the value value of ofthe the absorption absorption coefficient coefficient kabs kfromabs from Table Table 6. The6. The most most robust robust planar planar microwave microwave diode diode is DD30,is DD30, which which can withstand can withstand incident incident MW MW radiation radiation of ~1 of W ~1 without W without any any changes changes in inits its electricalelectrical parameters. parameters. The The diodes diodes DD0 DD0 and and DD1 DD155 were were ten-fold ten-fold weaker; weaker; their their characteris- characteristics ticsremain remain unchanged unchanged if if the the incident incident MW MW power power does does not not exceed exceed 100 100 mW. mW. A comparisonA comparison of ofthe the electrical electrical parameters parameters of ofseveral several millimeter millimeter wave wave zero zero voltage voltage biasedbiased room-temperature room-temperature diodes diodes is ispresented presented in in Table Table 7.7. Our planarplanar diodesdiodes areare not not inferior infe- riorto to the the other other diodes, diodes, inin the sense sense of of the the vo voltageltage responsivity responsivity at atlow low radiation radiation frequency. frequency. Moreover, though the voltage responsivity of the planar diodes decreases towards the higher frequencies, it remains sufficient to detect electromagnetic radiation in the sub- terahertz range. It should be emphasized, once again, that the planar diodes are attractive in their simplicity and do not require complex manufacturing technologies. Sensors 2021, 21, 4487 13 of 17
Table 7. Parameters of several different millimeter wave diodes operating at room temperature without bias voltage.
Dimension of R , R , Diode Technology R , kΩ v at 30 GHz v at 1 THz Reference Active Part 0 V/W V/W Planar DD15 LPE 1 µm 1.7 100 0.17 This paper Planar DD30 LPE 1 µm 14.3 160 0.33 This paper Bow-tie GaAs VPE 10 µm 4.0 0.3 0.1 [21] Gated bow-tie GaAs MBE 2 µm 5.5 4.0 2.0 [24] TEMF Gated bow-tie GaAs MBE 2 µm 5.5 70 0.06 [24] Schottky Heterojunction MBE 2 µm 1.0 300 0.4 [26] Bow-tie InGaAs MBE 12 µm 9.3 5.0 2.0 [36] Self-switching (SSD) MOCVD 90 nm 1.5 80 2.0 [37] Graphene ballistic Graphene/Boron 500 nm 6.0 800 0.05 [38] rectifier nitride
4. Conclusions Summarizing the results of the research of planar microwave sensors on the base of n-AlxGa1-xAs having different AlAs mole fractions (x = 0, 0.15 and 0.3), we can draw the following conclusions:
1. Voltage responsivity of the aluminum containing n-AlxGa1-xAs sensors is by two orders of magnitude higher than that of the n-GaAs-based diodes; 2. Higher voltage responsivity of the planar microwave diodes on the base of n-AlxGa1-xAs is related to effective rectification of the microwave currents on the quasi–ohmic contacts of the diodes; 3. Voltage responsivity of the planar microwave diodes is sensitive to the white-light illumination; the lower the AlAs mole fraction in the n-AlxGa1-xAs diode, the stronger its sensitivity to the light; 4. The characteristics of voltage responsivity versus temperature depend on the x content in the sensor material. As temperature increases, the responsivity of the n-GaAs-based diodes increases, while it decreases in the case of aluminum-containing diodes; the diode with a higher x value is more temperature-sensitive. Different behavior of the sensors with temperature is related to a different quality of the quasi–ohmic metal– semiconductor contacts of the diodes; a higher AlAs mole fraction in the compound semiconductor stipulates higher asymmetry of the diode’s IV characteristic. As a result, the diodes with a higher x value behave more like metal–semiconductor Schottky junction diodes; 5. The sensors on the base of the n-Al0.3Ga0.7As compound can withstand incident microwave radiation power up to 1 W.
Author Contributions: Conceptualization, A.S. and, S.A.; methodology, A.S., S.A., and J.G.; simula- tions, A.S., and A.C.,ˇ growth by LPE A.Š.; microwave investigation A.S., M.A., and A.L.; writing— original draft preparation, A.S.; writing—review and editing, S.A., and J.G.; visualization, A.S. and M.A.; supervision, A.S; project administration, J.G. All authors have read and agreed to the published version of the manuscript. Funding: This work was in part supported by the Research Council of Lithuania (grant No. S-LU-20-7). Acknowledgments: We kindly acknowledge Angele˙ Steikunien¯ e,˙ and Gytis Steikunas¯ for their assistance in sample preparation. Conflicts of Interest: The authors declare no conflict of interest. Sensors 2021, 21, x FOR PEER REVIEW 14 of 17
Acknowledgments: We kindly acknowledge Angelė Steikūnienė, and Gytis Steikūnas for their as- Sensors 2021, 21, 4487 sistance in sample preparation. 14 of 17 Conflicts of Interest: The authors declare no conflict of interest.
AppendixAppendix A A StatisticalStatistical presentation presentation of ofthe the voltage voltage respon responsivitysivity andand electrical electrical resistance resistance of the of pla- the narplanar microwave microwave diodes. diodes.
(a) (b)
(c) (d)
(e) (f)
FigureFigure A1. A1. Box-chartsBox-charts of of the the vo voltageltage responsivity responsivity ( (aa,c,c,e,e)) and and electrical electrical resistance resistance ( b,d,f) of the planarpla- nar microwave diodes DD0 (a,b); DD15 (c,d); DD30 (e,f)with different width of the small contact. microwave diodes DD0 (a,b); DD15 (c,d); DD30 (e,f)with different width of the small contact.
TheThe statistical statistical representation representation was executed byby testingtesting diodesdiodes in in batches batches each each containing contain- ing480 480 units. units. The The box-charts box-charts present present data data of of the the diodes diodes on on the the base base of of GaAs GaAs (diodes (diodes DD0), DD0), on onthe the base base of of Al Al0.150.15GaGa0.850.85AsAs (diodes(diodes DD15)DD15) andand onon thethe basebase ofof AlAl0.30.3GaGa0.70.7AsAs (diodes (diodes DD30) DD30) shown in Figure A1 (a, b), (c, d) and (e, f), respectively. Outputs of the fabricated planar diodes in the respective DD0, DD15 and DD30 batches are 83%; 78% and 46%.
Sensors 2021, 21, x FOR PEER REVIEW 15 of 17
Sensors 2021, 21, 4487 15 of 17 shown in Figure A1 (a, b), (c, d) and (e, f), respectively. Outputs of the fabricated planar diodes in the respective DD0, DD15 and DD30 batches are 83%; 78% and 46%.
(a) (b)
(c) (d)
(e) (f)
FigureFigure A2. A2. HistogramsHistograms of the of thevoltage voltage responsivity responsivity (a,c,e) ( aand,c,e )electrical and electrical resistance resistance (b,d,f) experi- (b,d,f) ex- mentalperimental data of data the of diodesw the diodesw DD0 (a DD0,b); DD15 (a,b); (c DD15,d); DD30 (c,d );(e DD30,f) approximated (e,f) approximated with Weibull with distribu- Weibull tion function. distribution function.
ExperimentalExperimental data data of of the the voltage voltage responsivity responsivity and and electrical electrical resistance resistance of of the the planar planar microwavemicrowave diodes diodes is is approximated approximated using using Weibull Weibull distribution function [[39]39] 𝛽 𝑥 " 𝑥 # 𝑓(𝑥)β= x β −1 exp − x β , (A1) f (x) = 𝜂 𝜂 exp − 𝜂 , (A1) η η η where x stands for either voltage responsivity or electrical resistance of the planar micro- wavewhere diode,x stands β and for η eithermark the voltage shape responsivity and the scale or parameters electrical resistanceof the Weibull of the distribution, planar mi- respectively.crowave diode, Theβ andhistogramsη mark theof the shape voltage and the re scalesponsivity parameters and electrical of the Weibull resistance distribution, of the planarrespectively. diodes Theare presented histograms in ofFigure the voltage A2. responsivity and electrical resistance of the planar diodes are presented in Figure A2.
Sensors 2021, 21, 4487 16 of 17
References 1. Goldsmith, P.F.; Hsieh, C.-T.; Huguenin, G.R.; Kapitzky, J.; Moore, E.L. Focal plane imaging systems for millimetre wavelengths. IEEE Trans. Microw. Theory Tech. 1993, 41, 1664–1675. [CrossRef] 2. Yujiri, L.; Shoucri, M.; Moffa, P. Passive millimeter wave imaging. IEEE Microw. Mag. 2003, 4, 39–50. [CrossRef] 3. Oka, S.; Togo, H.; Kukutsu, N.; Nagatsuma, T. Latest trends in millimeter-wave imaging technology. Prog. Electromagn. Res. Lett. 2008, 1, 197–204. Available online: http://www.jpier.org/PIERL/pierl01/24.07120604.pdf (accessed on 22 April 2021). [CrossRef] 4. Kancleris, Ž.; Laurinaviˇcius,A.; Anbinderis, T. Sensitivity of homogeneity mapping of semiconductor wafer using millimetre waves. Int. J. Infrared Millim. Waves 2004, 25, 1633–1644. [CrossRef] 5. Smith, B.T.; Weatherall, J.C.; Barber, J.B. Method for Identifying Materials Using Dielectric Properties through Active Millimeter Wave Illumination. US Patent US8946641B2, 6 April 2012. Available online: https://patents.google.com/patent/US8946641 (accessed on 22 April 2021). 6. Rappaport, T.S.; Sun, S.; Mayzus, R.; Zhao, H.; Azar, Y.; Wang, K.; Wong, G.N.; Schulz, J.K.; Samimi, M.; Gutierrez, F. Millimeter Wave Mobile Communications for 5G Cellular: It Will Work! IEEE Access 2013, 1, 335–348. [CrossRef] 7. Al-Ogaili, F.; Shubair, R.M. Millimeter-wave mobile communications for 5G: Challenges and opportunities. In Proceedings of the 2016 IEEE International Symposium on Antennas and Propagation, Fajardo, PR, USA, 26 June–1 July 2016. 8. Texas Instruments, Automotive mmWave Sensors. Available online: https://www.ti.com/sensors/mmwave-radar/automotive/ overview.html (accessed on 25 February 2021). 9. Texas Instruments, Automotive mmWave Sensors. Available online: https://www.ti.com/sensors/mmwave-radar/industrial/ overview.html (accessed on 25 February 2021). 10. Du, J.; Hellicar, A.D.; Li, L.; Hanham, S.M.; Nikolic, N.; Macfarlane, J.C.; Leslie, K.E. Terahertz imaging using a high-Tc superconducting Josephson junction detector. Supercond. Sci. Technol. 2008, 21, 125025. [CrossRef] 11. Xu, K.-X.; Essa, A.A.; Zhou, S.-P.; Bao, J.-S. Microwave detection using Y-Ba-Cu-O granular film bridge device. Phys. C Supercond. 2000, 339, 42–48. [CrossRef] 12. Tydex, Golay Cell GC-1D with Diamond Window. Available online: http://www.tydexoptics.com/products/thz_devices/golay_ cell/ (accessed on 5 March 2021). 13. Takahashi, T.; Sato, M.; Hirose, T.; Hara, N. GaAsSb-based backward diodes for highly sensitive millimetre-wave detectors. Electron. Lett. 2009, 45, 1269–1270. [CrossRef] 14. Cibiraite-Lukenskienˇ e,˙ D.; Ikamas, K.; Lisauskas, T.; Krozer, V.; Roskos, H.G.; Lisauskas, A. Passive Detection and Imaging of Human Body Radiation Using an Uncooled Field-Effect Transistor-Based THz Detector. Sensors 2020, 20, 4087. [CrossRef] [PubMed] 15. Liu, L.; Xu, H.; Duan, Y.; Lichtenberger, A.W.; Hesler, J.L.; Weikle, R.M., II. A 200GHz Schottky Diode Quasi-Optical Detector Based on Folded Dipole Antenna. In Proceedings of the 20th International Symposium on Space Terahertz Technology, Charlottesville, VA, USA, 20–22 April 2009. 16. Hesler, J.L.; Crowe, T.W. NEP and responsivity of THz zero-bias Schottky diode detectors. In Proceedings of the 32nd International Conference on Infrared, Millimeter and Terahertz Waves (IRMMW-THz), Cardiff, UK, 2–7 September 2007. 17. Hoefle, M.; Haehnsen, K.; Oprea, I.; Cojocari, O.; Penirschke, A.; Jakoby, R. Compact and Sensitive Millimetre Wave Detectors Based on Low Barrier Schottky Diodes on Impedance Matched Planar Antennas. J. Infrared Milli Terahz Waves 2014, 35, 891–908. [CrossRef] 18. Virginia Diodes, Inc. Zero Bias Detector. Available online: https://www.vadiodes.com/en/products/detectors (accessed on 1 March 2021). 19. Harrison, R.I.; Zucker, J. Microwave detection and mixing using the thermoelectric effect of hot carriers in semiconductors. Appl. Phys. Lett. 1963, 3, 153–154. [CrossRef] 20. Ašmontas, S.; Sužiedelis, A. New microwave detector. Int. J. Infrared Milli Waves 1994, 15, 525–538. [CrossRef] 21. Sužiedelis,˙ A.; Gradauskas, J.; Ašmontas, S.; Valušis, G.; Roskos, H.G. Giga- and terahertz frequency band detector based on an asymmetrically-necked n-n+-GaAs planar structure. J. Appl. Phys. 2003, 93, 3034–3038. [CrossRef] 22. Minkeviˇcius,L.; Tamošiunas,¯ V.; Kašalynas, I.; Seliuta, D.; Valušis, G.; Lisauskas, A.; Boppel, S.; Roskos, H.G.; Köhler, K. Terahertz heterodyne imaging with InGaAs-based bow-tie diodes. Appl. Phys. Lett. 2011, 99, 131101. [CrossRef] 23. Kašalynas, I.; Venckeviˇcius,R.; Valušis, G. Continuous Wave Spectroscopic Terahertz Imaging with InGaAs Bow-Tie Diodes at Room Temperature. IEEE Sens. J. 2013, 13, 50–54. [CrossRef] 24. Ašmontas, S.; Anbinderis, M.; Cerškus,ˇ A.; Gradauskas, J.; Sužiedelis,˙ A.; Šilenas,˙ A.; Širmulis, E.; Umansky, V. Gated bow-tie diode for microwave to sub-terahertz detection. Sensors 2020, 20, 829. [CrossRef] 25. Sužiedelis,˙ A.; Ašmontas, S.; Kazlauskaite,˙ V.; Gradauskas, J. Sensitivity increase of a point contact hot carrier microwave detector. Electron. Lett. 2009, 45, 1328–1329. [CrossRef] 26. Gradauskas, J.; Suziedelis, A.; Asmontas, S.; Sirmulis, E.; Kazlauskaite, V.; Lucun, A.; Vingelis, M. Sensitive planar semiconductor detector from microwave to infrared applications. IEEE Sens. J. 2010, 10, 662–667. [CrossRef] 27. Sužiedelis,˙ A.; Ašmontas, S.; Gradauskas, J.; Šilenas,˙ A.; Cerškus,ˇ A.; Luˇcun,A.; Paškeviˇc, C.;ˇ Anbinderis, M.; Žalys, O. Planar asymmetric dual diode for millimeter wave detection and power measurement. Lith. J. Phys. 2017, 57, 225–231. [CrossRef] Sensors 2021, 21, 4487 17 of 17
28. Sužiedelis,˙ A.; Ašmontas, S.; Gradauskas, J.; Šilenas,˙ A.; Luˇcun,A.; Cerškus,ˇ A.; Paškeviˇc, C.;ˇ Žalys, O.; Anbinderis, M. Pulsed microwave sensor on heavily doped semiconductor substrate. In Proceedings of the Progress in Electromagnetics Research Symposium (PIERS’17), Singapore, 19–22 November 2017; pp. 1037–1104. 29. Baca, A.G.; Ashby, C.I.H. Fabrication of GaAs Devices, 1st ed.; The Institution of Engineering and Technology: London, UK, 2005; p. 134. Available online: https://digital-library.theiet.org/content/books/cs/pbep006e (accessed on 28 April 2021). 30. Williams, R. Modern GaAs Processing Methods, 2nd ed.; Artech House: Norwood, MA, USA, 1990; 437p. 31. Mooney, P.M. Deep donor levels (DX centers) in III-V semiconductors. J. Appl. Phys. 1990, 67, R1–R26. [CrossRef] 32. Best, J.S. The Schottky barrier height of Au on n-Ga1−xAlxAs as a function of AlAs content. Appl. Phys. Lett. 1979, 34, 522–524. [CrossRef] 33. Lang, D.V.; Logan, R.A.; Jaros, M. Trapping characteristics and a donor-complex (DX) model for the persistent-photoconductivity trapping center in Te-doped AlxGa1-xAs. Phys. Rev. B 1979, 1015–1030. [CrossRef] 34. Rhoderick, E.H.; Williams, R.H. Metal-Semiconductor Contacts, 2nd ed.; Clarendon Press: Oxford, UK, 1988; 252p. 35. Harrison, R.G. Detector and Mixers. In Microwave Solid State Circuit Design, 2nd ed.; Bahl, I., Bhartia, P., Eds.; John Wiley & Sons Inc.: Hoboken, NJ, USA, 2003; pp. 559–627. 36. Seliuta, D.; Kašalynas, I.; Tamošiunas,¯ V.; Balakauskas, S.; Martunas,¯ Z.; Ašmontas, S.; Valušis, G.; Lisauskas, A.; Roskos, H.G.; Köhler, K. Silicon lens-coupled bow-tie InGaAs-based broadband terahertz sensor operating at room temperature. Electron. Lett. 2006, 42, 825–826. [CrossRef] 37. Sangaré, P.; Ducournau, G.; Grimbert, B.; Brandli, V.; Faucher, M.; Christophe Gaquière, C.H.; Ana Íñiguez-de-la-Torre, A.; Íñiguez- de-la-Torre, I.; Millithaler, J.F.; Mateos, J.; et al. Experimental demonstration of direct terahertz detection at room-temperature in AlGaN/GaN asymmetric nanochannels. J. Appl. Phys. 2013, 113.[CrossRef] 38. But, D.B.; Consejo, C.; Coquillat, D.; Teppe, F.; Knap, W.; Varani, L.; Torres, J.; Auton, G.; Zhang, J.; Hill, E.; et al. Millimeter and submillimeter range detector based on Graphene Ballistic Rectifiers. In Proceedings of the 22nd International Microwave and Radar Conference (MIKON), Pozna´n,Poland, 15–17 May 2018; pp. 585–588. [CrossRef] 39. Characteristics of the Weibull Distribution. Available online: https://www.weibull.com/hotwire/issue14/relbasics14.htm (accessed on 6 May 2021).