We present the synthesis of the organic-based ferrimagnet vanadium tetracyanoethylene (V[TCNE]x, x~2) via low temperature chemical vapor deposition (CVD). This optimized recipe yields an increase in Curie temperature from 400 K to over 600 K and a dramatic improvement in magnetic resonance properties.
Recent progress in the field of organic materials has yielded devices such as organic light emitting diodes (OLEDs) which have advantages not found in traditional materials, including low cost and mechanical flexibility. In a similar vein, it would be advantageous to expand the use of organics into high frequency electronics and spin-based electronics. This work presents a synthetic process for the growth of thin films of the room temperature organic ferrimagnet, vanadium tetracyanoethylene (V[TCNE]x, x~2) by low temperature chemical vapor deposition (CVD). The thin film is grown at <60 °C, and can accommodate a wide variety of substrates including, but not limited to, silicon, glass, Teflon and flexible substrates. The conformal deposition is conducive to pre-patterned and three-dimensional structures as well. Additionally this technique can yield films with thicknesses ranging from 30 nm to several microns. Recent progress in optimization of film growth creates a film whose qualities, such as higher Curie temperature (600 K), improved magnetic homogeneity, and narrow ferromagnetic resonance line-width (1.5 G) show promise for a variety of applications in spintronics and microwave electronics.
The organic-based ferrimagnetic semiconductor vanadium tetracyanoethylene (V[TCNE]x, x~2) exhibits room temperature magnetic ordering and promises the advantages of organic materials for magnetoelectronic applications, such as flexibility, low cost production, and chemical tunability. Previous studies have demonstrated functionality in spintronic devices, including hybrid organic/inorganic1,2 and all-organic spin valves3, and as a spin polarizer in an active organic/inorganic semiconductor heterostructure4. In addition, V[TCNE]x~2 has demonstrated promise for inclusion in high frequency electronics due to its extremely narrow ferromagnetic resonance linewidth5.
There are four different methods which have been established for synthesizing V[TCNE]x~26-9. V[TCNE]x~2 was first synthesized as powder in dichloromethane via reaction of TCNE and V(C6H6)6. These powders exhibited the first room temperature magnetic ordering observed in an organic-based material. However, the powder form of this material is extremely air sensitive, limiting its application in thin film devices. In 2000, a chemical vapor deposition (CVD) method was established for creating V[TCNE]x~2 thin films7. More recently physical vapor deposition (PVD)8 and molecular layer deposition (MLD)9 have also been used to fabricate thin films. The PVD method requires an ultra-high vacuum (UHV) system and both PVD and mLD methods require extremely long times to grow films thicker than 100 nm, whereas the CVD films can easily be deposited in thicknesses ranging from 30 nm to several microns. In addition to the variety of thicknesses available with the CVD method, extensive studies have yielded optimized films that consistently show high quality magnetic properties including: narrow ferromagnetic resonance (FMR) linewidth (1.5 G), high Curie temperature (600 K), and sharp magnetic switching5.
Magnetic ordering in V[TCNE]x~2 thin films proceeds via an unconventional route. SQUID magnetometry measurements show strong local magnetic ordering, but the absence of X-ray diffraction peaks and featureless transmission electron microscopy (TEM)10 morphology reveal a lack of long-range structural order. However, extended X-ray absorption fine-structure (EXAFS) studies11 show that each vanadium ion is octahedrally coordinated with six different TCNE molecules, indicating a robust local structural order with a vanadium-nitrogen bond length of 2.084(5) Å. Magnetism arises from an antiferromagnetic exchange coupling between the unpaired spins of the TCNE– radical anions, which are distributed across the entire TCNE– molecule , and the spins on the V2+ ions, leading to a local ferrimagnetic ordering with TC ~ 600 K for optimized films5. In addition to exhibiting room temperature magnetic ordering, V[TCNE]x~2 films are semiconducting with 0.5 eV bandgap12. Other properties of note include possible sperimagnetism below a freezing temperature of ~150 K13,14, anomalous positive magnetoresistance12,15,16, and photo-induced magnetism13,17,18.
The CVD method for synthesizing V[TCNE]x~2 thin films is compatible with a variety of substrates due to low temperature (<60 °C) and conformal deposition. Previous studies have shown successful deposition of V[TCNE]x~2 on both rigid and flexible substrates7. Further, this deposition technique lends itself to tuning through modification of precursors and growth parameters.19-22 While the protocol shown here yields the most optimized films to date, significant progress has been made in improving some of the film properties since the discovery of this method and further gains may be possible.
1. Synthesis and Preparation of Precursors
2. Set up Deposition System inside an Argon Glovebox
3. Clean up
Figure 1. (A) Fully assembled custom chemical vapor deposition (CVD) system. (B) Expanded view of the components for the CVD system. Please click here to view a larger version of this figure.
Figure 2. (A) A top view of the substrates in the reactor showing their location. (B) Approximate film thickness as function of position inside the reactor tube, Part A from Figure 1B for a deposition of 75 min. Please click here to view a larger version of this figure.
The first and easiest method for determining if a deposition is successful is to do a visual inspection of the films. The film should appear dark purple with a mirror finish that is uniform across the substrates. If there are spots on the surface of the substrate where there is no V[TCNE]x~2 or it is lighter in color, then this is likely due to the presence of solvents or other impurities on the substrate surface. Additionally the film should be opaque. Unless a thin film was deposited over a short time span of only a few minutes, translucent films often means there may have been a problem with the flow rate of the precursors during the deposition.
It is important to note that in addition to sub-optimal growth conditions, atmospheric exposure can degrade the film which may result in films whose qualities appear to be less favorable; therefore it is essential to prevent oxygen exposure when transporting and measuring the samples for analysis. Transportation of the sample outside the glovebox requires encapsulation of the film with materials such as epoxy24 or parylene25 or enclosing the sample in custom-designed cans that fit the measurement tool4. Local structure and film composition can be characterized by X-ray photoemission spectroscopy (XPS) and Fourier transform infrared spectroscopy (FTIR).
Magnetic properties can be measured using a SQUID magnetometer. Optimized films yield an extrapolated Curie temperature (TC) around 500-600 K. Due to film breakdown above RT, the value of TC is extracted from a magnetization versus temperature measurement, such as the one shown in Figure 3A. This measurement is performed in a Quantum Design SQUID Magnetometer with an applied field of 100 Oe. The presence of a large splitting of the zero-field cooled (ZFC) and field-cooled (FC) magnetization values at low temperature is evidence of isolation of local spin environments and is a larger presence in lower quality films. The TC of the films can be extracted by fitting the magnetization values above the peak to the Bloch law
Ms(T) = Ms(0)(1 – BT3/2),
where MS is the saturation magnetization and B is a fitting parameter. For the data shown in Figure 3A this fit yields a TC of 600 K.
In addition to characterizing the magnetic response to temperature, the magnetization as a function of applied field can also be measured resulting in a hysteresis loop like the one shown in Figure 3B. For optimized films the switching of the magnetization is sharp, achieving saturation by 100 Oe. The coercivity should be about 20 Oe at 300 K.
Ferromagnetic resonance (FMR) studies are a key technique for identifying successful film growth. The presence of a single, narrow peak in the FMR measurement is strong evidence of an ideal growth. The best films have full width at half maximum (FWHM) linewidth on the order of 1-2 G. Measurements of resonance sub-optimal growth will result in a spectrum which shows multiple resonance features at some or all angles of rotation. Figure 4 shows the FMR spectrum of an ideal film at various angles of the applied microwave and DC fields, rotating from in-plane (90°) to out-of-plane (0°) at 300 K with an applied microwave frequency of 9.85 GHz. The samples are normalized to account for variation in the magnitude of the intensity due to cavity conditions.
Electrical properties of the films can be characterized through transport measurements. The simplest measurement geometry is a two-probe measurement to measure current as a function of voltage for various temperatures. Figure 5A shows a film deposited on glass with 30 nm of Al and 40 nm of Au top contacts created by thermal evaporation. Electrical contact is made through indium press to a custom air-tight puck for a Quantum Design physical properties measurement system (PPMS). Current-voltage (I-V) measurements are performed using a Keithley 2400 sourcemeter. These measurements reveal Ohmic I-V characteristics at all temperatures with resistance that increases with decreasing temperature as shown in Figure 5B. The temperature dependent resistance data can be fit to an Arrhenius equation
R = R0 e-Ea/kBT,
to extract an activation energy, Ea~0.50 eV. This value represents the band gap energy in the electronic structure for this semiconducting material12.
Figure 3. (A) Field cooled (open circles) and zero field cooled (filled circles) magnetization versus temperature with an applied magnetic field of 100 Oe. Solid black line is a fit used to extract TC of 600 K. (B) Magnetization versus field measured at 300 K. Please click here to view a larger version of this figure.
Figure 4. Room temperature FMR spectra as a function of angle from in-plane (90°) to out-of-plane (0°). Please click here to view a larger version of this figure.
Figure 5. (A) Schematic of sample structure for transport sample. (B) Resistance values extracted from the current-voltage measurements shown in the inset for temperatures from 150 K to 300 K. Please click here to view a larger version of this figure.
The key parameters for V[TCNE]x~2 deposition include temperature, carrier gas flow, pressure, and ratio of precursors. Because the chemical vapor deposition set-up is not commercially available these parameters will need to be optimized for each system. A previous study by Shima et al. revealed that the temperature has the largest impact on the sublimation rate of the TCNE precursor26. The temperature can be modified both by the value set on the temperature controller and also by making adjustments to the wire spacing on the heating coil and as such will need to be calibrated for each system. Temperature calibration is performed by measuring inside the reactor before fully assembling the system for deposition. It is important to place the TCNE boat in the hottest zone of the reactor at a temperature near 75 °C.
The next most important parameter is carrier gas flow. The carrier gas flow rate for the TCNE should be higher than for V(CO)6. The recommended flow rates are 56 sccm for the V(CO)6 and to 84 sccm for the TCNE, and it is important to monitor these flow rates during deposition to ensure stability (a sampling frequency of roughly 10 min is typically sufficient).
If the pressure is above 35 mmHg the reaction will likely not occur. If the pressure is high and the reaction has not started (there is no V[TCNE]x~2 appearing) there is likely a leak in the system. A large leak means the system will not pump down at all, but if there is a small leak the system may reach 40-50 mmHg. The first place to check for leaks is at all of the glassware connections. Most commonly, the vacuum grease on the flow lines can get dirty and needs to be wiped clean and replaced. In addition to leaks, pressure problems can be caused by unclean glassware or the presence of contaminants that outgas inside the chamber. For this reason it is important to carefully consider any material placed within the reaction chamber.
In addition to optimizing the reaction parameters, the surface treatment of substrates is critical for good film growth. V[TCNE]x~2 can be deposited onto a wide variety of substrates, but the surface must be clean and free of residual solvents. Even touching substrate surfaces with tweezers can contaminate them. Also, samples that have been processed may require additional cleaning steps. For example, to deposit V[TCNE]x~2 onto photoresist, the photoresist must have been baked for long enough to remove any trace of solvents. Additionally, for deposition of V[TCNE]x~2 onto a chemically processed surface, such as a self-assembled monolayer may require semiconductor grade chemicals for the processing.
CVD-grown films of V[TCNE]x~2 are ideal for incorporation in device structures; however there is limited processing that can be done to the V[TCNE]x~2 films because they are sensitive to solvents, water, air, and high temperatures. V[TCNE]x~2 films can be shadow masked for thermal, e-beam, or sputter deposition of other organics or metals. Various encapsulation techniques can be used to transport samples with V[TCNE]x~2 to measurement tools, but is a challenge for working with this material. However, this difficulty is also common for other organic devices, such as organic light emitting diodes (OLEDs), so there is a significant body of work on techniques for encapsulation27-29.
Beyond the ability to grow films of V[TCNE]x~2 for many different applications, this method of chemical vapor deposition is suitable to chemical tunability and exploration of other types of organic thin films, such as V[MeTCEC]30. This technique provides the ability to create a thin film organic magnet in a thickness ranging from a few tens of nanometers to several micrometers for applications from spintronics devices to microwave applications and beyond.
The authors have nothing to disclose.
This work was supported by NSF Grant No. DMR-1207243, the NSF MRSEC program (DMR-0820414), DOE Grant No. DE-FG02-03ER46054, and the OSU-Institute for Materials Research. The authors acknowledge the NanoSystems Laboratory at Ohio State University, and technical assistance from C. Y. Kao and C.Y. Chen.
Equipment | |||
Nitrogen Glovebox | Vacuum Atmospheres | Omni | steps done in nitrogen glovebox can also be done in an argon glovebox |
1 L three-neck round bottom flask | Corning | 4965A-1L | |
500 mL round bottom flask | Sigma Aldrich | 64678 | |
Turbo vacuum pumping station | Agilent Varian | G8701A-011-037 | |
Glass Stopcock | Kontes | 185000-2440 | |
Glass two way connecting tube | Corning | 8940-24 | Corning Pyrex(R) 105 degree Angled Tube Adapter with Two-Way 24/40 Standard Taper Joint |
Coldfinger | Custom part made by OSU chemistry glass shop | ||
Argon Glovebox | Vacuum Atmospheres | Nexus I | |
Hot plate stirrer | Corning | 6795 | |
Thermoeletric cooler | Advanced Thermoelectric | TCP-50 | |
Temperature controller | Advanced Thermoelectric | TLZ10 | for TE cooler |
Power supply | Advanced Thermoelectric | PS-145W-12V | for TE cooler and temperature controller |
Temperature controller | J-Kem Scientific | Model 150 | For heating coil |
Heating wire | Pelican Wire Company | Nichrome 60 | |
Custom glassware pieces | Made by OSU Chemistry glass shop | ||
Vacuum pump | BOC Edwards | XDS-5 | Connected to the CVD set-up |
Flow meter | Gilmont | GF-2260 | |
Micrometer valve | Gilmont | 7300 | Controls flow of argon over TCNE |
Micrometer valve | Gilmont | 7100 | Controls flow of argon over V(CO)6 |
Tubing | Tygon | R3603 | 1/8 in walls, connected between valves and meter |
3-way Stopcock | Nalgene | 6470 | used to adjust the flow rates |
Pressure gauge | Matheson | 63-4105 | connects to the top of Figure 1 part A |
SQUID magnetometer | Quantum Design | MPMS-XL | |
EPR | Bruker | Elexsys | |
PPMS | Quantum Design | 14T PPMS | |
Sourcemeter | Keithely | 2400 | |
Materials | |||
Sodium metal | Sigma Aldrich | 262714 | |
Anthracene | Sigma Aldrich | 141062 | |
Anhydrous tetrahydrofuran | Sigma Aldrich | 186562 | |
Vanadium(III) chloride tetrahydrofuran complex | Sigma Aldrich | 395382 | |
Carbon monoxide gas | OSU stores | 98610 | |
Tetraethylammonium bromide | Sigma Aldrich | 241059 | |
Phosphoric acid | Sigma Aldrich | 79622 | |
Methanol | Sigma Aldrich | 14262 | |
Silcone oil | Sigma Aldrich | 146153 | |
Copper pellets | Cut from spare copper wire | ||
Tetracyanoethylene | Sigma Aldrich | T8809 | |
Glass slides | Gold Seal | 3010 | |
Activated Charcoal | Sigma Aldrich | 242276 |