1887
banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Invited Review Article: A 10 mK scanning probe microscopy facility
Rent:
Rent this article for
USD
10.1063/1.3520482
/content/aip/journal/rsi/81/12/10.1063/1.3520482
http://aip.metastore.ingenta.com/content/aip/journal/rsi/81/12/10.1063/1.3520482
View: Figures

Figures

Image of FIG. 1.
FIG. 1.

Plot of the figure of merit for cryogenic SPM systems, , vs year. Various sub-K SPM systems reported over the last 30 years are listed along with their locations. Red dots refer to UHV compatible systems, while systems represented by blue dots have low vacuum (often cryo-vacuum) only. References: IBM Watson, 1987 (Ref. 19); UC Berkeley, 1999 (Ref. 21); Dartmouth College, 1999 (Ref. 33); TU Delft, 2001 (Ref. 35); IBM Almaden, 2004 (Ref. 65); Universitat Hamburg, 2004 (Ref. 25); Laboratory for Physical Sciences, 2004 (Ref. 38); RIKEN, 2006 (Ref. 27); ETH Zurich, 2007 (Ref. 39); University of Tokyo, 2007 (Ref. 40); Tsinghua University, 2007 (Ref. 28); Leiden University, 2008 (Ref. 30); McGill University, 2008 (Ref. 41); University of Tokyo, 2009 (Ref. 31); Universidad Autónoma de Madrid, 2010 (Ref. 42). The dot labeled NIST refers to the system described in this article.

Image of FIG. 2.
FIG. 2.

3D computer aided design (CAD) model overview of the processing lab in the ULTSPM facility at NIST. The lab hosts MBE growth chambers for metal (1) and for III-V semiconductor samples (2), an RF furnace (3), an FIM chamber for tip preparation and analysis (4), a stand-alone quick access cryogenic STM/AFM (5) and a two-stage load lock chamber (6a,b). Horizontal transfer chambers called “bidirectional translator stations” (7) connect each chamber to the central interlab vacuum line (8). Part of this line (9) is portable and can be disconnected from the main system to decouple from the processing lab. This section (which has its own ion pump) can be moved to connect with the UHV chamber in the ULTSPM lab (Fig. 4). The portable part reaches through two hatches in the walls of the double shielded room to make the connection.

Image of FIG. 3.
FIG. 3.

3D CAD model of a bidirectional translator station. Two rotational feedthroughs (blue and red) on the main 254 mm flange (1) control, respectively, the two-way motion of the rail (2) and the sled (3) on top of the rail. A 4-slot tip/sample holder carousel (4) is mounted on top of the sled. In the interlab vacuum line a fixed rail (5) supports a 0.61 m (24 in.) long sled (6). The sled has two posts with magnetic blocks (7) so it can be dragged along by a handheld magnet (8) on the outside of the vacuum system. This sled has two 4-slot carousels.

Image of FIG. 4.
FIG. 4.

3D CAD model cross section of the double shielded ULTSPM lab. The inner room is built upon a separate ≈110 t concrete slab (1) supported by six actively controlled pneumatic isolators. A ≈6 t granite table (2), also with active pneumatic isolators, supports the measurement equipment. The ≈1 t cryostat (3), passively isolated, is mounted in a hole in the granite table and in the concrete slab. Inside the cryostat the dilution refrigerator insert (4) hangs directly immersed in the liquid helium bath. Above the cryostat the UHV chamber (5) links to the central interlab vacuum line (6) that provides access to the processing lab, which is located to the right of this image. The UHV chamber, if disconnected from the central interlab vacuum line, can be moved sideways to the left on a pair of rails (7) to allow the DR insert to be removed. A long vertical welded bellows translator (8) rotates inside a centering ring (11) and is used to move the SPM module into and out of the cryostat. The entire lab is shielded by an inner (9) and an outer (10) acoustic enclosure, the inner enclosure also acting as an RF shield. Two hatches in the walls (on the right) can be opened to connect the system to the processing lab.

Image of FIG. 5.
FIG. 5.

Translator bayonet coupling mechanism and radiation-blocking tube shutter design. (a) A photograph of a tube shutter held by a translator. (b) 3D CAD model of the tube shutter assembly: (1) top cover, (2) side wings, (3) side post, (4) coil spring, (5) tungsten wire, and (6) bottom cover. (c)–(d) Cross sectional views of the shutter operation in a tube. (c) Retracting a push-pin (gray colored arrow), the side wings extend out by the force of the internal coil spring, and push against the tube wall, fixing the shutter solidly in the tube. (d) To release and pick up the shutter, a center push-rod pushes into the center of the shutter to retract the side wings with the tube shutter supported on the translator anchors. In this way the tube shutter can be positioned in the DR central tube at any desired height to block thermal radiation.

Image of FIG. 6.
FIG. 6.

Photograph and 3D CAD models of the SPM module and parts. (a) Front view photograph of the SPM module. (b) 3D CAD models showing parts of the SPM module unit. (c) 3D CAD model of the SPM tip holder and its receptacle. (d) 3D CAD model of the SPM sample holder and its receptacle. Label details: (1) a tube shutter module used to clamp the SPM module, (2) compression stage with inner compression spring, (3) rotation stage for pin alignment, (4) piezoelectric motors for XY-coarse motion, (5) receptacles for the SPM tip and a sample (details are shown in (c) and (d), (6) front cover with a plate spring of the Z-motor, (7) piezoelectric motors for Z-coarse motion, (8) capacitance sensor to measure Z-coarse motion, (9) bottom cover including all signal pins, (10) slot for wobble stick operation, (11-a) and (11-b) three bottom and two top signal contacts for both the SPM tip holder and sample holder, (12) center hole for in situ evaporation and optical access, (13) five contact points to be used between device leads and signal contacts (11), (14) spring clip for kinematic and electric contact, (15) ground plane.

Image of FIG. 7.
FIG. 7.

Characteristics of the Z-piezo motor driven by the NIST piezo motor controller. (a) Step size vs. spring tension using a 210 V drive-voltage amplitude. The spring tension for the Z-motor was adjusted in situ by rotating one of the four screws holding the clamping plate using a UHV wobble stick. (b) Step size vs. signal amplitude after optimizing the motor tension. The solid lines are linear fits with slopes, 1.31 and 1.64 nm V−1, for the forward (blue) and backward motion (red), respectively. (c) Calibration data for the capacitance sensor inside of the SPM module. The capacitance change scales linearly with piezo motor motion with a slope of 0.37 aF nm−1. (d) Capacitance sensor measurements of the motor step size vs drive-voltage amplitude. The solid lines are linear fits with slopes, 0.43 and 0.51 aF V−1, for the forward (blue) and backward motion (red), respectively. All measurements in (a)–(d) were made in UHV at room temperature. The step sizes and distances were measured using using an optical telescope on a theodolite. The step sizes were determined by measuring the number of steps to move the motor 1 mm. Errors bars in distance measurements represent one standard deviation in uncertainty in the determination of the distance traveled using the optical theodolite.

Image of FIG. 8.
FIG. 8.

Phase diagram of the He–He mixture. Below 870 mK (the triple point) the mixture undergoes a spontaneous phase separation into a He-rich phase and a He-dilute phase. At T = 0 K, the He-rich phase approaches 100% He, but the He-dilute phase maintains a finite He concentration of 6.4%. Adapted from Ref. 85.

Image of FIG. 9.
FIG. 9.

Schematic drawing of the NIST dilution refrigerator system. There are five heat exchangers: E1 for the 1 K pot, E2 for the still, E3 for the JT condenser, E4 a continuous heat exchanger, and E5 a sintered silver heat exchanger. Z1 is the main impedance for the 1 K pot condensing mode and Z3 is the impedance for the JT condensing mode. These two condensing modes can be switched by two independent valves (colored yellow and orange). Z2 is a secondary impedance to prevent a backdraft of He. Two needle valves for the 1 K are shown, which can be controlled manually (pink colored) or automatically (purple colored). The JT return line includes a compressor (labeled Comp) for high flow operation.

Image of FIG. 10.
FIG. 10.

Enthalpy vs pressure diagram for He with simplified closed cycle paths for the three different condensing modes used in the DR (see text). (a) Cycle for the 1 K pot condensing mode. (b) Cycle for the JT condensing mode with a low return pressure and flow rate. (c) Cycle for the high pressure JT condensing mode using a compressor for high flow operation. Figure adapted from Ref. 90.

Image of FIG. 11.
FIG. 11.

3D CAD model of the cryostat and DR insert. (a) 3D CAD model of the LHe cryostat and UHV compatible dilution refrigerator with the 15 T superconducting magnet system. (b) Magnified view of the dilution refrigerator showing the main components (red box in (a)). (c) and (d) show the closed and open position of the plate shutter on the still plate used to block thermal radiation. Similar plate shutters are located on the IVC main flange and ICP plate.

Image of FIG. 12.
FIG. 12.

A photograph of the bottom portion of the DR with the extension and receptacle for the SPM module.

Image of FIG. 13.
FIG. 13.

Thermometry calibration and cooling measurements. (a) Susceptibility of the CMN thermometer measured at the superconducting fixed points: Ir (96 mK), AuAl (161 mK), AuIn (208 mK), Cd (502 mK), and In(3300 mK). The different symbols show the dependence on excitation amplitude and frequency. (b) Calibration of mixing chamber RX thermometer sensor against the CMN measured temperature. (c) CMN and RX sensor measurements during a JT condensing and cooling cycle from 4 K. The oscillations at high temperature are due to successive mixture additions and pumping cycles.

Image of FIG. 14.
FIG. 14.

DR cooling power as a function of mixing chamber temperature for different circulation flow rates using the 1 K pot condenser. Cooling power is measured by applying a fixed heater power to the mixing chamber and recording the equilibrium temperature. (a) Symbols show the measured data and solid lines show the maximum cooling power predicted by Eq. (4). (b) Solid lines show the cooling power given by Eq. (5) with a that varies as a function of circulation flow rate (see Fig. 16).

Image of FIG. 15.
FIG. 15.

DR cooling power as a function of mixing chamber temperature for different circulation flow rates using the Joule-Thomson condenser. Cooling power is measured by applying a fixed heater power to the mixing chamber and recording the equilibrium temperature. (a) Symbols show the measured data and solid lines show the maximum cooling power predicted by Eq. (4). (b) Solid lines show the cooling power given by Eq. (5) with a loss that varies as a function of circulation flow rate (see Fig. 16).

Image of FIG. 16.
FIG. 16.

The heat loss as a function of mixture flow rate from fitting the 1 K pot (blue circles) and JT (maroon squares) cooling power data in Figs. 14 and 15 to Eq. (5). The solid line is a fit to the combined data sets with a slope of W per μmol s−1 and an intercept of 0.61 μW.

Image of FIG. 17.
FIG. 17.

Equilibrium mixing chamber temperature versus circulation flow rate for different heating powers applied to the mixing chamber for (a) the 1 K pot condenser mode, and (b) the JT condenser mode. For the 1 K pot mode a minimum temperature of 9.1 mK occurs at 300 μmol s−1, and for the JT mode the minimum temperature of 9.4 mK occurs at 350 μmol s−1.

Image of FIG. 18.
FIG. 18.

Vibration spectra at the SPM stage obtained with both (a) horizontal and (b) vertical geophones. These results were measured while running the dilution refrigerator in either of two condensing modes, 1 K pot and JT, and various flow rates. A vibration spectrum without the DR mixture gas circulating (after recovering mixture into dumps) is also shown for comparison.

Image of FIG. 19.
FIG. 19.

3D CAD model of the three vibration isolation stages. (a) Stage 1 (green) consists of a 110 t concrete mass with a double keel design, supported by six pneumatic air springs (black). The size of the inner hole is 1.42 m × 1.42 m. (b) Stage 2 (red) consists of a 6 t granite table supported by four pneumatic air springs (black). The size of the inner hole for the cryostat is 1.27 m × 1.27 m. (c) Stage 3 (blue) consists of an aluminum ring suspended by four stainless steel rods and containing four pneumatic air springs.

Image of FIG. 20.
FIG. 20.

3D CAD model of the He–He mixture recovery pumping line. Three gimbal systems (Ref. 86) (1) decouple the recovery pumping line at each vibration isolation stage from the SPM system. In between the gimbal systems, the pumping line is rigidly anchored to the building [(2), grey], the first vibration isolation stage [(3), green], the second stage [(4), red] and the third stage [(5), blue]. At (6) the line passes through (and anchors to) the wall of the inner shielded room, and at (7) it passes through (and anchors to) the outer shielded room.

Image of FIG. 21.
FIG. 21.

Vibration transmission spectra of the three-stage isolation system in Figs. 4 and 19. (a) Theoretical transmission spectrum for the combined three-stage isolation system. (b-d) Transmission spectra measured between (b) floor and stage 1, (c) stage 1 and stage 2, and (d) stage 2 and stage 3.

Image of FIG. 22.
FIG. 22.

Characteristics of the three-stage vibration isolation system. (a) Spectral density (SD) measurements of the velocity vs frequency of stage 1 (red), 2 (blue), and 3 (green) with all stages floating. Velocity spectra were measured with an accelerometer (Ref. 112) in velocity mode. SD measurements of (b) the tunneling current and (c) the Z-height signal vs frequency with stage 3 only or stage 2 and stage 3 floating. These measurement was performed on a Ag(111) sample with a W tip by using the external preamplifier (Ref. 113) and a spectrum analyzer (Ref. 114). Tunneling parameters: tunneling current setpoint 100 pA, sample bias 2.0 V. DR parameters: 100 μmol s−1 in JT mode, T = 13 mK.

Image of FIG. 23.
FIG. 23.

Block diagram showing the analog (solid line) and digital (dotted line) signal connections and the electronics for the ULTSPM system. The light gray colored boxes are NIST-built electronics. A feedback circuit and a scan controller are integrated into a single servo box. Thick gray outlines indicate the RF shielded room and the UHV part of the dilution refrigerator (SPM module inside). All of the signal cables (except for an optical fiber connected to a precision lock-in amplifier) pass through RF filters, which are fixed to the wall of the RF shielded room.

Image of FIG. 24.
FIG. 24.

Piezo-motor drive circuitry and performance measurments. (a) Drive signals originating from the DAC are amplified × 30 and sent to the six 2.5 nF piezo stacks in the SPM module (1), via six independently fired Triac switches (2). The switches are timed by a digital clock, and sequentially-fired. An additional 100 nF capacitance (3) acts as a charge reservoir to allow quick piezo charging. (b) Full amplitude output signal driving a piezo capacitance load. (c) Expanded time scale for all six transition signals (slip mode) driving a piezo capacitance load. The time delay between each signal is ≈28 μsec, and the transition time from 90% to 10% of the signal height is ≈872 ns.

Image of FIG. 25.
FIG. 25.

Cryogenic current preamplifier concepts and schematics. (a) Schematic diagram of an STM tunneling unit and preamplifier. stands for a stray capacitance coming from signal wiring inside the STM chamber to the external preamplifier. Symbols and are the input current and voltage noise of the preamplifier. The tunneling junction (a dotted oval) can be expressed with an equivalent circuit diagram (left-most diagram). (b) The frequency response of the preamplifier and noise. (c) A preamplifier circuit diagram using a pair of silicon p-channel MOSFETs, where parts inside the dotted region are at cryogenic temperature. As indicated in the text the p-channel silicon MOS devices were replaced by GaAs devices shown in (d). (d) A preamplifier circuit diagram using a pair GaAs MESFETs, where parts inside the dotted region are attached to the still plate in the DR (≈700 mK). and are ordinarily to . and are n-channel dual gate transistors (Ref. 122) and is an operational amplifier (Ref. 123). The symbol I stands for a constant-current source, whose value can be determined by the operating point of the transistor (−5 to −200 μA) in the present circuit. and are the feedback resistor (, , or ) and the bias voltage, respectively. A relay is used to switch between an external preamplifier and this internal preamplifier.

Image of FIG. 26.
FIG. 26.

Effects of the piezo motor operation and magnetic field ramping on the SPM measurements. (a) Temperature of the mixing chamber, sample holder, and SPM receptacle from RX sensors vs time during the operation of the Z-piezo motor. The first rise in temperature comes from walking backward by 20 steps with 230 V, and the second while approaching forward by 6 coarse and 4 fine steps with 280 and 240 V, respectively. (b) SPM servo Z-height signal vs time while the magnetic field was changed from 13.5 to 13.25 T with a ramping speed of 0.1 T min−1. The feedback loop was closed during the measurement. Point (1) corresponds to the opening of the superconducting switch from persistent mode, (2) the beginning of the field ramp, and (3) the end of the field ramp. The fine oscillations during the ramp correspond to atomic corrugation due to raster scanning an image while the field was ramping.

Image of FIG. 27.
FIG. 27.

Spectral density (SD) measurements of the current vs frequency measured on Ag (111) with a W tip using an external preamplifier (Ref. 113) and a NIST-built internal preamplifier [Fig. 25(d)]. Data is shown for cases of an open feedback loop (red and orange), a closed feedback loop (blue and light blue), and for the tip retracted out of tunnel range (black and gray). (a) Frequency band DC to 3.2 kHz and (b) Frequency band DC to 100 Hz. Tunneling parameters: tunneling current setpoint 100 pA, sample bias 2.0 V. DR parameters: JT mode flow rate 100 , T = 13 mK.

Image of FIG. 28.
FIG. 28.

Spectral density (SD) of the current noise vs frequency, in both 1 K pot and JT condensing modes, with different mixture flow rates of 100 and 200 μmol s−1. (a) Frequencies from DC to 3.2 kHz and (b) Frequencies from DC to 100 Hz. All the measurements were done with an open servo feedback loop to be more sensitive to external mechanical noise. Green curves indicate the same measurements with the tip retracted, which determines the baseline (zero tunnel current) electronic noise. Data were measured on Ag (111) surface with a W tip using the NIST-built internal preamplifier [Fig. 25(d)] and a spectrum analyzer. (Ref. 114) Tunneling parameters: tunneling current setpoint 100 pA, sample bias 2.0 V, T = 13 mK.

Image of FIG. 29.
FIG. 29.

Spectral density (SD) of the Z-height signal in both the 1 K pot and JT condensing modes with different mixture flow rates of 100 and 200 μmol s−1. (a) DC to 3.2 kHz frequency band and (b) DC to 100 Hz frequency band. Green curves are the measurement with an open feedback loop, which determines the base noise limit of the electronics at fixed Z. Data were measured on a Ag (111) surface with a W tip using the NIST-built internal preamplifier [Fig. 25(d)] and a spectrum analyzer (Ref. 114). Tunneling parameters: tunneling current setpoint 100 pA, sample bias 2.0 V, T = 13 mK.

Image of FIG. 30.
FIG. 30.

STM images of single Er atoms deposited onto a CuN surface grown on a Cu(100) single crystal. (a) Topographic STM image with the Z-height shown in a gray scale. Image size is 8 nm × 8 nm. (b) 3D rendered color version of the image in (a) which highlights the surface corrugation. The blue arrow points to a single adatom. The CuN regions (corrugated areas) are separated by regions of the bare Cu(100) surface regions (smooth areas). Tunneling parameters: tunneling current setpoint 200 pA, sample bias 10 mV. T = 13 mK

Image of FIG. 31.
FIG. 31.

STM measurements of epitaxial graphene grown on the C-face SiC. (a) High resolution STM image of the epitaxial graphene sample measured at 13 mK. Image size 5 nm × 5 nm. Tunneling parameters: tunneling current setpoint 100 pA, sample bias −250 mV. (b) Z-height profile cross-section taken along the white line in (a).

Image of FIG. 32.
FIG. 32.

Spectroscopic STS measurements of epitaxial graphene. (a) dI/dV spectrum showing Landau levels at B = 2 T. The N = 0 Landau level is observed at −134 mV with over 20 other Landau levels clearly resolved. Tunneling parameters: tunneling current setpoint 300 pA, sample bias −300 mV, modulation voltage 1 mV. (b) High resolution dI/dV spectrum of the N = 1 Landau level at B = 11.5 T showing the lifting of the four-fold degeneracy due to electron spin (up and down arrows) and valley (blue and red) degeneracies. Tunneling parameters: tunneling current setpoint 400 pA, sample bias −250 mV, modulation voltage 50 μV.

Loading

Article metrics loading...

/content/aip/journal/rsi/81/12/10.1063/1.3520482
2010-12-29
2014-04-20
Loading

Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Invited Review Article: A 10 mK scanning probe microscopy facility
http://aip.metastore.ingenta.com/content/aip/journal/rsi/81/12/10.1063/1.3520482
10.1063/1.3520482
SEARCH_EXPAND_ITEM