Virginia Polytechnic Institute and State University, Department of Chemistry
Blacksburg, VA 24061
Four chlorinated phenyltricyanoethylenes were synthesized and reacted with vanadium hexacarbonyl to form ferrimagnetic coordination polymers. Species chlorinated at the 2, 3, or 4 position of the ring as well as a dichlorinated analogue with substitution at both the 2 and 6 positions were examined. The magnetic ordering temperatures of the polymers were reproducible and ranged from 140 to 288 K. All compounds displayed soft ferrimagnetic behavior, and their structure-property relationships are discussed.
Keywords: Magnetic coordination polymers, phenyltricyanoethylene, ferrimagnetism
In 1991, Manriquez and coworkers reported the discovery of the first room temperature molecule-based magnet, prepared from the reaction of vanadium hexacarbonyl and tetracyanoethylene (TCNE). This compound, believed to be a network of V2+ bridged by TCNE radical anions, is limited in application due to its rapid decomposition upon exposure to air. A possible source of the air-sensitivity is the V2+ ions, as these are highly susceptible to oxidation. Replacing V2+ with later transition metal ions such as Mn2+ or Fe2+ may overcome this limitation, however, such substituted compounds have been found to be non-magnetic at room temperature. Therefore, discovering a replacement for TCNE has become desirable for synthesizing useful, room temperature molecule-based magnets. The goal of this study was to find an acceptor starting material, which when paired with one of these later transition metals, would yield an air-stable, room temperature magnet.
However, as only vanadium was used in this study, it was hoped that a way to maximize the temperature at which the material is magnetic through tuning only the acceptor could be found. Once this was known, different metals could be utilized to find air-insensitive room temperature magnets.
2.1 Magnetic Ordering
Given a network of species containing unpaired electrons, there are two general types of magnetic interaction that can lead to long-range magnetic order with a net magnetic moment. One type, termed ferromagnetism, occurs when all of the electron spins are aligned in the same direction. Conversely, antiferromagnetism occurs when adjacent spins in a compound align in opposite directions, effectively canceling each other out and resulting in no net magnetic moment. Ferrimagnetic ordering is similar to antiferromagnetic ordering, only adjacent spins are unequal in magnitude resulting in a net magnetic moment. Figure 1 illustrates ferromagnetic, antiferromagnetic, and ferrimagnetic ordering schematically.
In most compounds, measurements show the net magnetization diminishing at elevated temperatures as enough kinetic energy becomes available to overcome the thermodynamic tendency to align at lower temperatures. This is because increased thermal energy allows electrons to be excited to higher orbitals, which are not necessarily aligned with the applied magnetic field. The temperature below which the moments start ordering is referred to as the Curie temperature (Tc).
2.2 Molecule-based Magnets
Two different types of molecule-based magnet starting materials were used in this project. The first is an easily reduced organic molecule (aka acceptor), that is designed to bridge two or more metal ions upon accepting an electron from each. The second starting material consisted of vanadium atoms (derived from vanadium hexacarbonyl), which contribute electrons to two organic molecules to form the divalent V2+ cation in conjunction with the loss of its six carbonyl functional groups. The remaining (three) d-electrons on the cation give rise to local moments which do not completely cancel each other out, resulting in a net magnetic moment. Magnets created in this manner are of technological importance because their magnetic properties can be tuned by varying either of the two starting materials (i.e. using slightly different molecules as the acceptor, or by using different metal ions).
Molecule-based magnets can be advantageous over atom-based magnets. For example, molecule-based magnets can often be synthesized at or near ambient conditions as opposed to their atom-based counterparts, which typically require very high temperatures for formation. In addition, molecule-based magnetic materials often have a higher magnetization per unit mass ratio than atom-based magnets. The cause of this is two fold: first, lighter nonmetal elements are often incorporated into the atomic structures, and second, the molecule-based magnetic compounds tend to have an open-framework type structure as opposed to the dense, tightly packed atomic lattices of atom-based magnets.
Although these types of compounds are amorphous, the overall stoichiometry for V[TCNE]x•y[CH2Cl2] type materials has been determined to be roughly two electron acceptor molecules for every vanadium ion through elemental analysis (x ~ 2). As illustrated in Figure 2, the result is one unpaired electron (a net spin of ½) per formula unit.
This study attempted to determine whether magnets with metals ions other than vanadium could be synthesized which retain the high ordering temperature of vanadium-based magnets, while being immune to decomposition upon exposure to air. To test this hypothesis, the research was conducted with the specific purpose of probing how the electronic structure of the organic bridges connecting the metal atoms affects the ordering temperature. The various organic bridges synthesized for study are shown in Figure 3. These organic bridges are all similar to TCNE but feature a phenyl ring that may be substituted with chlorine to modify the steric (size) and electronic properties of the molecule, which, when reacted, allows for tunability of the synthesized magnets.
3.1 General Considerations
Magnet synthesis was performed in a chemically inert nitrogen atmosphere, while all other procedures were carried out under ambient conditions. Vanadium hexacarbonyl was synthesized following procedures outlined in the literature.
Solvents used for synthesis of the magnetic polymers and electrochemistry were distilled from P2O5. All other solvents were used as received. 2-(2-chlorophenyl), 2; 2-(3-chlorophenyl), 3; 2-(4-chlorophenyl), 4; and 2-(2,6-dichlorophenyl)-1,1,2tricyanoethylene,5 were synthesized in a similar manner to 2-Phenyl-1,1,2-tricyanoethylene,1; as shown in Figure 4.
First, 2,6-dichlorobenzaldehyde, 2.27 g (1.3 mmol), was added to 20 mL 100 % ethanol in a 100 mL beaker. Next, 0.874 g of 1.32 mmol malononitrile was added to the ethanol/benzaldehyde solution. Then, three drops of piperidine were added, and the solution was stirred briefly. The solution was then placed in an evaporating dish to remove the solvent. The collected solid was then redissolved in dichloromethane and stirred with decolorizing carbon. The resulting solution was filtered into a separate flask, and the remaining solvent was evaporated using a rotoevaporator.
First, 1.76 g (7.55 mmol) of  (synthesized as shown above) was added to 200 mL of 100% ethanol in a 250 mL beaker. 1.01 g of 15.3 mmol potassium cyanide was dissolved in 50 mL of deionized water contained in a separate 500 mL beaker. Once both solids had dissolved, the ethanol solution was added to the aqueous solution. Next, 250 mL of deionized water was then added to the aqueous solution to double the solution volume. Then, the reaction vessel was placed in an ice bath for 45 minutes, after which, the vessel was removed from the ice bath and 2 mL of chilled, concentrated hydrochloric acid was added. Next, the solution was stirred briefly, and then the vessel was covered overnight. Finally, the solid formed was collected via vacuum filtration and a buchner funnel.
First, 0.450 g (1.799 mmol) of  (synthesized as shown above) was added to a 100 mL of diethylether, while 0.5 g of Nchlorosuccinimide were added to 80 mL of deionized water in a separate 100 mL beaker. Once the solids in both solutions had dissolved, the aqueous solution was added to the ether solution. Next, the resulting mixture was stirred vigorously for 15 min. Then, the ether layer was collected with a separatory funnel and dried with several scoops of anhydrous sodium sulfate. Next, the solution was filtered into a new flask, and the ether was evaporated using a rotoevaporator. The collected solid was then purified by column chromatography using silica gel as the stationary phase and chloroform (CHCl3) as the elutant.
Magnetic Solid, 
First, 37.4 mg, 0.171 mmol, of vanadium hexacarbonyl were dissolved in 1 mL degassed dichloromethane (CH2Cl2) in a 10 mL sample vial, while 90.8 mg of 0.366 mmol of  were added to 2 mL degassed CH2Cl2 in a 50 mL round bottom flask. Next, the vanadium solution was added to the acceptor solution by filtering through a coarse glass frit. Then, the vessel was sealed, and the solution stirred with a magnetic stirbar for approximately 30 minutes. Finally, the resulting black solid was then collected by filtering through a medium fritted glass filter and dried in-vacuo for 1 hour.
3.4 Material Characterization
A CH Instruments Model 600A Potentiostat was used for electrochemical analysis. Specimens were prepared with 0.125 mol of a given ethylene solid added to 25 mL of a 0.1 M [n-Bu4N][PF6] solution of acetonitrile. The electrochemical potential of the sample was then scanned from 0 to -800 mV at a rate of 100 mV/s. A polished carbon working electrode was used with a Ag/AgCl reference electrode in a 10 mL glass cell. All magnetic measurements were collected with a Quantum Design MPMS XL Superconducting Quantum Interference Device (SQUID). Curie temperatures of the magnets were found by cooling from 300 to 5 K in a magnetic field of 10 Oe and measuring the magnetization of the sample from 5 to 300 K in 1 K increments. Magnetic hysteresis measurements were performed at 5 K with a scanning range of -100 to 100 Oe in increments of 3 Oe.
Infrared (IR) spectra for the compounds were obtained as potassium bromide (KBr) pellets by adding the compound to powdered KBr. All spectra were obtained in the range of 4000 - 400 cm-1 on a MIDAC M-series FTIR spectrometer.
All 1H NMR spectra were acquired in CDCl3 with the use of a Varian Inova400 spectrometer.
4. Results and Discussion
Table 1 gives the relevant yield and characterization data for the syntheses reported in section 3.3.
Nuclear magnetic resonance (NMR) and IR techniques were chosen as the primary methods for characterizing the products of our reactions. As can be seen for , a singlet at 7.95 ppm results from the benzyllic proton and the protons on the aromatic ring give rise to the multiplet at 7.45 ppm. Analysis of  shows no major difference in the chemical shift of the aromatic protons but the introduction of two doublets. These doublets disappear from the spectrum after the reaction with NCS, giving evidence that the Hydrogen atoms are eliminated as a double bond is formed in . Similar behavior was observed in the other acceptors.
NMR analysis of the magnetic solid  was not performed due to the insolubility of the polymer. IR spectra give evidence towards the formation of the polymer, as the CN stretch moves to lower wavenumbers, which is presumably from electron donation by the vanadium ions into the antibonding orbitals of the acceptor. The other magnetic polymers synthesized displayed similar behavior.
4.2 Critical Temperature
As can be seen in Figure 5, the Tc is affected by the position of substitution on the arene ring. The unsubstituted acceptor gives a polymer with a Tc near 200 K, while the 2, 3, and 2,6 substituted acceptors gives polymers with Tc's of 240, 260, and 288 K, respectively. Surprisingly, substitution at the 4 position of the phenyl ring results in a magnetic material that orders at a lower temperature than the unsubstituted material. Substitution at the 2 or 3 positions each cause ordering above that of the unsubstituted acceptor, and substitution at both the 2 and 6 positions on the ring gives an enhanced Tc raising effect which leads to a polymer with the highest Tc out of all those studied in this paper. The measurements of the Tc for each of the polymers have been reproducible within ± 5 K.
Figure 5. Critical temperature data for the synthesized
Figure 6. Magnetic hysteresis curve for the 2,6-dichlorotri-
cyanoethylene coordination polymer
At first glance, this trend can be rationalized by recognizing the π-electron donating ability of the chlorine atom. It appears that at the 2 and 3 positions on the ring, the σ-withdrawing effect of the chlorine atom is more powerful than the π-donating effect. At the 4 position, however, the donating effect could overcome the withdrawing effect, resulting in a lower measured Tc. By donating electron density into the arene ring, which is already electron rich, communication between the spins of the vanadium ions is lessened, which correlates to a lower ordering temperature. However, electrochemical measurements presented later contradict this approach.
4.3 Magnetic Hysteresis
The coercivity (Hc) is the value of the applied field where the curve crosses the x-axis. The saturation magnetization is the maximum value reached by the magnetization. All compounds exhibited coercivities (that is, the magnitude of the field required to return the magnetization of a material back to zero) between 1 and 5 Oe, with the exception of the polymer made from substitution at the 4 position, which required 11 Oe. Table 2 lists the Hc values as they correspond to each polymer.
All of the compounds are soft magnets by definition, as hard magnetic materials show Hc values in excess of 125 Oe. The measured Hc values stem from the low anisotropy of the vanadium atoms. A system shows small Hc values when the starting metal ion used is essentially isotropic, which is expected for these compounds as the three unpaired electrons from octahedral vanadium would reside in the t2g orbitals.
The saturation magnetization of these compounds also provides insight into their magnetic coupling. One mole of unpaired electrons contributes 5585 emu*G/mol to the magnetization of a material. Therefore, the saturation magnetization for the polymer formed with the 2,6-dichlorinated acceptor should be in excess of 26000 emu*G/mol for ferromagnetic interactions (5 unpaired electrons × 5585 emu*G/mol). If we assume a 2:1 stoichiometric ratio of electron acceptors to vanadium ions, then the saturation magnetization is calculated to be 4923 emu*G/ mol, which is closer to the net magnetization resulting from one mole of unpaired electrons, indicating that the synthesized compounds are ferrimagnets.
4.4 Cyclic Voltammetry
Table 3 gives important information concerning our electron acceptors. The potentials displayed directly correlate to the voltage output required by the potentiostat to either reduce (shown in the middle column) or oxidize (shown in the far right column) the sample. The middle column shows that reducing the acceptor molecules becomes easier when chlorine is substituted onto the phenyl ring. This contrasts with what is observed for the Tc data, where the Tc of the polymer made with position 4 is lower than that of the polymer made with position 1. This difference between the two data sets suggests that some factor other than simply the electron withdrawing/donating ability of chlorine is affecting the T.
Figure 7 shows the cyclic voltammogram of the unsubstituted phenyltricyanoethylene. All of the compounds were shown to be reversible by cyclic voltammetry, which suggests that the bridging electron acceptors are stable upon accepting an electron from the vanadium during synthesis.
Changing the position of the halide functional group on the phenyl ring altered the Tc of the synthesized magnet, giving rise to tunability. Substitution at the 4 position of the phenyl ring in phenyltricyanoethylene produced a solid upon reaction with V(CO)6. These solids ordered at lower temperatures than the polymers made with unsubstituted acceptors. Substitution at the 2 and 3 positions yielded polymers with higher Tc than the polymer made with the unsubstituted acceptor. From this data, it appears that the Tc increasing effect from chlorine substitution is additive, as the Tc of the compound created from the disubstituted acceptor was higher than for any of the other compounds, which means that a potential pathway towards reaching a maximum Thas been found for this class of compounds. Further studies would allow us to explore this route and hopefully arrive at higher Tc magnetic materials than were reported here.
The magnetic polymers synthesized were found to be air sensitive. The limited applications for air-sensitive molecule-based magnets warrant future work to synthesize polymers using other non-vanadium first row transition metals. In addition, synthesizing magnets made from poly-substituted acceptors may hold promise for improved room-temperature magnets. If the Tc increasing effects of chlorine substitution are truly additive, synthesizing a magnet with 2, 3, 5, 6-tetrachlorinated phenyltricyanoethylene would display the highest possible Tc for this class of magnetic compounds.
We thank Virginia Tech Analytical Services for the use of their FTIR and NMR equipment, Professor Mark Anderson for the use of the CH Instruments potentiostat, Mark Harvey for the supply of vanadium hexacarbonyl, and the Petroleum Research Fund administered by the American Research Fund for funding this reserach. This research was performed under the direction of Professor Gordon Yee.
 Manriquez, J. M.; Yee, G. T.; McLean, R. S.; Epstein, A. J.; Miller, J. S. A room-temperature molecular/organicbased magnet. Science 1991, 252, 1415.
 Zhang, J.; Ensling, J.; Ksenofontov, V.; Gutlich, P.; Epstein, A. J.; Miller, J. S. [MII(TCNE)2]×xCH2Cl2 (M = Mn, Fe, Co, Ni) molecule-based magnets with Tc values above 100 K and coercive fields up to 6500 Oe. Angewandte Chemie International Edition 1998, 37(5), 657-660.
 McQuaid, M.J.; Gole, James L. The effect of carbonyl complexation on highly exothermic vanadium oxidation reactions. Chemical Physics 2000, 260, 367-382.
 Vickers, Elaine B.; Senesi, Andrew; Miller, Joel S. Ni[TCNE]2•zCH2Cl2 (Tc = 13 K) and VxNi1-x[TCNE]y•zCH2Cl2 solid solution room temperature magnets. Inorganica Chemica Acta 2004, 357, 3889-3894.
 Kaul, B. B.; Yee, G. T. Two new acceptor building blocks for 'high Tc' coordination polymer magnets. Inorganica Chemica Acta 2001, 326, 9-12.
 Jiles, D. Chapter 4. Introduction to Magnetism and Magnetic Materials. Chapman & Hall: London, 1991.
 Haskel, D.; Islam, Z.; Lang, J.; Kmety, C.; Srajer, G.; Pokhodnya, K. I.; Epstein, A. J.; Miller, J. S. Local structural order in the disordered vanadium tetracyanoethylene room-temperature molecule-based magnet. Physical Review B 2004, 70(5), 054422.
 Torre, Edward D. Chapter 1. Magnetic Hysteresis. Institute of Electrical and Electronics Engineers, Inc.: New York, 1999.
About the Author
Joseph M. Zadrozny
Joe Zadrozny is a fifth year undergraduate at Virginia Tech. His major is chemistry and he is currently completing a minor in mathematics. After he finishes this year, he will be going to graduate school. His current top choice schools are mostly on the west coast where the weather is consistently great (note: he is not even considering going anywhere near Seattle). He enjoys reading and cycling in his spare time, and sometimes he has even been known to fold origami.