A Quantitative Determination of Magnetic Nanoparticle Separation Using On-Off Field Operation of Quadrupole Magnetic Field-Flow Fractionation (QMgFFF)

Abstract

Quadrupole Magnetic Field-Flow Fractionation (QMgFFF) is a technique for characterization of sub-micrometer magnetic particles based on their retention in the magnetic field from flowing suspensions. Different magnetic field strengths and volumetric flow rates were tested using on-off field application and two commercial nanoparticle preparations that significantly differed in their retention parameter, λ (by nearly 8-fold). The fractograms showed a regular pattern of higher retention (98.6% v. 53.3%) for the larger particle (200 nm v. 90 nm) at the higher flow rate (0.05 mL/min v. 0.01 mL/min) at the highest magnetic field (0.52 T), as expected because of its lower retention parameter. The significance of this approach is a demonstration of a system that is simpler in operation than a programmed field QMgFFF in applications to particle mixtures consisting of two distinct particle fractions. This approach could be useful for detection of unwanted particulate contaminants, especially important in industrial and biomedical applications.

Introduction

In recent years, there has been a growing interest in applying magnetic nanoparticles in ever widening fields, such magnetic fluids,¹ catalysis,^2,3 biotechnology and biomedicine,⁴ magnetic resonance imaging,^5,6 data storage,⁷ and environmental remediation.^8,9 For many of these applications, the characterization of the magnetic nanoparticles is of key importance because their optical and magnetic properties depend strongly on their chemical composition, structure and physical dimensions.^10–12 A major difficulty in analyzing the magnetic nanoparticle in aqueous suspensions is lack of control over local magnetic and flow forces acting on particulate analytes. To alleviate this problem, approaches based on mechanical vibration to the column,¹³ intermittent magnetic field,¹⁴ increasing shear forces exerted on the analysis sample,¹⁵ and differential magnetic catch and release,¹⁶ have been proposed, with varying degrees of success.

This study focused on field on-off operation of a QMgFFF system. The technique belongs to a large family of the field-flow fractionation (FFF) analytical techniques,^17–18 here specialized for the characterization of magnetic micro- and nano-particles.¹⁹ The effort in the past was directed towards a programmed field operation to measure the mass of the magnetic species per microparticle and its distribution in the sample.^19–21 The purpose of this study was to evaluate the system using a simpler on-off field operation for potential future applications requiring removal of contaminating weakly magnetic material prior to structural and magnetization analysis, such as by small angle X-ray scattering.²²

Experimental

Materials

All materials were of analytical grade and were used as they were received from the manufacturer without further purification. Polydisperse dextran-coated magnetite nanoparticles were tested. Particle A had a nominal mean diameter of 200 nm (catalog No. 18150, StemCell Technologies Inc., Vancouver, Canada). The smaller Particle B had a nominal diameter range of 20 – 100 nm (Anti Phycoerythrin MicroBeads, catalog no. 120-000-294, Miltenyi Biotec GmbH, Bergisch Gladbach, Germany). The particle size distribution in suspension was measured experimentally by dynamic light scattering (Nicomp 380 ZLS, Port Richey, FL). Particle A had a mode of 200 nm and polydispersity index of 0.132; the same parameters for Particle B were 90 nm and 0.49, respectively. The particle volume magnetic susceptibility in SI units system was measured by us for similar nanoparticles and by the same manufacturers but in different batches: the mean susceptibility evaluated at a mean B of 1.27 T for Particle A was 0.0011(3) and for Particle B was 0.000035(12), where the number in parentheses indicates precision.²⁴ Biodegradable FL-70 detergent was obtained from Thermo Fisher Scientific, Inc., USA, added to Milli-Q water (0.1% v/v), then degassed and filtered using a 0.22 µm pore Millipore filter prior to use as the main component of the mobile phase (carrier solution) and for sample dilution (described below).

Apparatus

The QMgFFF system has been described previously.^19–21 Briefly, it consisted of a thin, ribbon-like helical slot (overall length of 23.5 cm, a depth that defined the flow channel thickness of w = 250 µm, and a breadth of about 1.6 cm) machined into a Delrin (Du Pont) rod fitted into a stainless steel cylinder (0.51 mm wall, 15.88 mm O.D.), resulting in a calculated channel volume of 0.94 mL. It was in liquid contact with an injection valve (7725i, Rheodyne, Cotati, CA), a mobile phase pump (515 HPLC pump, Waters Corp., Milford, MA), and an optical detector (fixed wavelength 254 nm, HyperQuan, Inc., Colorado Springs, CO) . The QMgFFF system schematic is shown in Supporting Information (Fig. S1).

The QMgFFF electromagnet (aperture 1.6 cm, length 15.24 cm) was powered by a Xantrex HPD60-5 regulated DC power supply (Xantrex Technology Inc., British Columbia, Canada), which is capable of generating a field at the pole tips in excess of the maximum of 0.52 T used in this study. The field was monitored using a model 6010 gauss meter (F. W. Bell, Orlando, FL) (Figs. S2 and S3, Supporting Information). The FFF theory is governed by the retention parameter λ = kT/F_mw that measures compaction of the separand at the accumulation wall.²¹ It is independent of the mobile phase viscosity and is inversely proportional to the magnetic force, F_m, which is proportional to the mass of the magnetic material incorporated in the particle (where k is the Boltzmann constant, T the absolute temperature in kelvins and w is the channel thickness). Thus, the retention parameter λ decreases with the mass of the magnetic material incorporated in the particle. Comparing 90 and 200 nm particles, the latter having over eight times the overall volume but similar in composition, will likely result in a retention parameter λ that is eight times lower and their increased retention. Therefore, considering the large difference between the retention parameters of the two nanoparticle preparations used in this study, it is expected that under suitable flow rate conditions the nanoparticles with higher magnetic mass will elute only after the field is turned off.^20,21

Procedure

The stock magnetic particle suspension was diluted 1:4 with carrier solution. A 20 µL volume of sample was introduced to the channel under various conditions of field intensity (0, 0.1, 0.21 and 0.52 T, corresponding to 0, 12.5, 26.25 and 65 T/m field gradients) and volumetric flow rate (0.05, 0.1, 0.5 and 1.0 mL/min). The magnetic field was first switched on prior to sample injection and then off when the elution peak trace returned to baseline (typically, within tens of minutes) as the flow conditions remained constant.

Results and Discussion

The polydisperse Particle A and Particle B samples showed distinct elution patterns (referred to as “fractograms”) upon injection into the QMgFFF system, as expected.²⁰ To evaluate non-specific particle retention in the system, the sample was injected at zero magnetic field (Particle A, Fig. 1). Indeed, the elution profiles for different flow rates were comparable, indicating no non-specific particle loss (further confirmed by lack of any residual particle elution after increasing the flow rate to a maximum of 1 mL/min; not shown). The noticeable drop of the peak value at the low flow rate of 0.05 mL/min was likely due to impaired particle elution at low local viscous shear stresses.

Fig. 1.

Control fractograms of Particle A nanoparticle suspension at 0 T applied field and different volumetric flow rates of the mobile phase.

For the test runs, the magnetic field was adjusted to a desired target value before sample injection. After apparent elution of the portion of injected material indicated by the light absorption signal returning to baseline, the magnetic field was set to zero, and as the flow continued, a second peak followed. The retention yield Y, defined as the proportion of the mass of the magnetically retained particles to the total mass of the sample, was calculated with the following equation:

$$Y=\frac{\mathit{\text{Area Peak}}2}{\mathit{\text{Area Peak}}1+\mathit{\text{Area Peak}}2}\times 100\%$$ (1)

where Area Peak 1 and Area Peak 2 are areas under the curves recorded with the field on and the field off, respectively.

Figure 2 summarizes retention results (fractograms) obtained for the larger of the two particles (Particle A) for the specific case of the field at 0.1 T. The highest retention yield (Y = 98.6%) was achieved for the larger particle when the nanoparticles were injected at the lowest flow rate of 0.05 mL/min with the highest applied magnetic field of 0.52 T (Fig. 3). Overall, the retention was a monotonic function of the field intensity and flow rate, such that at 0.05 mL/min and 0.21 T it was slightly lower than that for the above mentioned conditions, and at 1.0 mL/min and 0.1 T it decreased to Y = 10.0%. The smaller nanoparticles (Particle B) were retained at a markedly smaller yield (Y = 53.3%, Fig. 4) even at the most favorable conditions (0.52 T and 0.01 mL/min). A linear extrapolation of the magnetic field values to 100% capture efficiency resulted for Particle A in 0.53 T for the lowest flow rate of 0.05 mL/min and 0.70 T for the highest flow rate of 1.0 mL/min; the respective values for Particle B were: 1.08 T for 0.01 mL/min and 1.70 T for 0.1 mL/min.

Fig. 2.

Fractograms of Particle A nanoparticle suspension at 0.1 T applied field and different flow rates of the mobile phase. The applied field was turned off at the end of elution of the first peak. A complete retention, Y = 100%, would correspond to the area of Peak 1 equal to zero.

Fig. 3.

Summary of mean magnetic retention yields, Y, obtained for Particle A suspension for different applied magnetic fields and different mobile phase flow rates. Each measurement was done in triplicate, and the relative standard deviations did not exceed 0.022. The retention was nearly complete (98.6%) for the highest field and the lowest flow rate

Fig. 4.

Summary of mean magnetic Particle B retention yields, Y, for different magnetic fields and different mobile phase flow rates. Each measurement was done in triplicate, and the relative standard deviations did not exceed 0.029. The maximum retention, Y = 53.3%, was markedly lower than that for Particle A for comparable field and flow conditions. No retention was measured at the lowest field of 0.1 T (unlike that for Particle A, compare with Fig. 3).

The ratio of elution volumes of a non-retained material to a retained material is known as a retention ratio, R, which is a non-linear function of the retention parameter, λ, namely R = 6λ[coth(1/2λ)-2λ]. (A thorough discussion of the relationship between parameters R and λ, and their use in interpreting fractograms is provided in references.^19–21) Referring to Fig. 2, the retention ratio of the material eluted in the second peak is less than 0.5 taking into account that none of it was eluted at 1 mL and that the second peak mode is at 1.5 mL (note that the mobile phase flow rate does not enter into the definition of λ). In comparison, the retention ratio of the material eluted in the first peak is very close to unity as it is almost indistinguishable from the non-retained peak (note irregular shape of the first peak suggesting overlap). Thus from Fig. 2, it can be seen that the retention ratio, R, of the material eluted in the second peak is less than half of that for the first peak, resulting in the ratio of their respective retention parameters, λ, to be less than 0.11 as calculated from the formula shown in the text above. This agrees well with the predicted ratio of 1/8 = 0.125 based on the size differences between Particle A and Particle B described in section Apparatus above.

Conclusions

The on-off field operation of the QMgFFF system allows quantification of the mass fractions of the retained and unretained portions of a magnetic nanoparticle preparation for given field and flow conditions. It could be used as a means of analytical-scale separation of weakly magnetic contaminants that interfere with magnetic and structural analysis of the nanoparticle suspensions.