Rui C.R. Rodrigues

Streamlined, two-column, simulated countercurrent chromatography for binary separation

We report on a numerical and experimental study of two-column versions of streamlined, multicolumn, semi-continuous chromatography for binary separation. The systems combine a flexible node design, cyclic flow-rate modulation, and relayed operation of the inlet/outlet ports to extend the mass-transfer zone over the largest possible length, while keeping it inside the system at all times. One advantage of these streamlined designs is the simplicity of their physical realization: regardless of the number of columns, they only require two pumps to supply feed and desorbent into the system, while the flow rates of liquid withdrawn from the system are controlled by material balance using simple two-way valves. In one case, an extra pump is needed to recirculate the fluid in closed-loop. A rigorous model-based optimization approach is employed in the optimal design of the cycles to generate solutions that are physically realizable in the experimental set-ups. The optimized schemes for two-column operation supply fresh feed into the middle of the system where the composition of the circulating fluid is closest to that of the feedstock fluid, and recover the purified products, extract and raffinate, alternately at the downstream end of the unit, while desorbent is continuously supplied into the upstream end of the system. By internally recycling part of the non-pure cut fraction, the scheme with a step of closed-loop recycling significantly reduces its solvent consumption. The feasibility and effectiveness of the reported two-column processes have been verified experimentally on the linear separation of nucleosides by reversed phase subject to 99% purity constraints on both products. It is shown that our processes compare favorably against single-column batch chromatography, steady-state recycling, and four-column, open-loop SMB, for the same amount of adsorbent; they are also better than the four-column, closed-loop SMB at high feed throughputs.

Chiral separation by two-column, semi-continuous, open-loop simulated moving-bed chromatography.

A two-column version of a multicolumn, semi-continuous, open-loop chromatograph for chiral separation is presented and validated experimentally. The heart of the process is a flexible node design and cyclic flow-rate modulation that succeed at keeping the mass-transfer zone inside the system without resorting to any recycling technique. One advantage of this streamlined design is the simplicity of its physical realization: regardless of the number of columns, it only requires two pumps to supply feed and desorbent into the system, while the flow rates of liquid withdrawn from the system are controlled by material balance using simple two-way valves. A rigorous model-based optimization approach is employed in the optimal cycle design to generate a solution that is physically realizable in the experimental apparatus. The optimized scheme for two-column operation supplies fresh feed into the system where the composition of the circulating fluid is closest to that of the feedstock fluid, and recovers the purified products, extract and raffinate, alternately at the downstream end of the unit while desorbent is supplied into the upstream end of the system. The feasibility and effectiveness of the two-column process are verified experimentally on the separation of reboxetine racemate, a norepinephrine re-uptake inhibitor, under overloaded conditions. Our set-up employs an automated on-line enantiomeric analysis system, comprising an analytical HPLC set-up with two UV detectors to monitor the composition profile at the downstream end of one of the columns; this monitoring system does not use a polarimeter.

A new multicolumn, open-loop process for center-cut separation by solvent-gradient chromatography.

A comprehensive description of a new process--the GSSR (Gradient with Steady State Recycle) process--for center-cut separation by solvent-gradient chromatography is provided, highlighting its versatility, flexibility, and ease of operation. The GSSR process is particularly suited for ternary separation of bioproducts: it provides three main fractions or cuts, with a target product contained in the intermediate fraction. The process comprises a multicolumn, open-loop system, with cyclic steady state operation, that simulates a solvent gradient moving countercurrently with respect to the solid phase. However, the feed is always injected into the same column and the product always collected from the same column as in a batch process; moreover, both steps occur only once per cycle. The GSSR process was experimentally validated in a pilot unit, using the purification of a crude peptide mixture by reversed phase as a proof of concept; the crude mixture is roughly 50% pure and some of its impurities have isocratic retention times very close to that of the target peptide. Experimental results are reported in terms of cyclic steady-state profiles and process performance indicators, which include product purity and yield. A simplified model-based approach, which uses only a few key components of the crude mixture, is employed to assist in the explanation of the process operation. By dynamically adjusting the switching interval while the process is running, to correctly position the composition profile with respect to the outlet ports, pure product satisfying the target specifications--98% purity and 95% recovery--was obtained under stable operation in the pilot unit.

Determination of competitive isotherms of enantiomers by a hybrid inverse method using overloaded band profiles and the periodic state of the simulated moving-bed process.

A procedure for determination of adsorption isotherms in simulated moving-bed (SMB) chromatography is presented. The parameters of a prescribed adsorption isotherm model and rate constants are derived using a hybrid inverse method, which incorporates overloaded band profiles of the racemic mixture and breakthrough data from a single frontal experiment. The latter are included to reduce the uncertainty on the estimated saturation capacity, due to the dilution of the chromatograms with respect to the injected concentrations. The adsorption isotherm model is coupled with an axially dispersed flow model with finite mass-transfer rate to describe the experimental band profiles. The numerical constants of the isotherm model are tuned so that the calculated and measured band profiles match as much as possible. The accuracy of the isotherm model is then checked against the cyclic steady state (CSS) of the target SMB process, which is readily and cheaply obtained experimentally on a single-column set-up. This experiment is as expensive and time consuming as just a few breakthrough experiments. If necessary, the isotherm parameters are adjusted by applying the inverse method to the experimental CSS concentration profile. The method is successfully applied to determine the adsorption isotherms of Trögers base enantiomers on Chiralpak AD/methanol system. The results indicate that the proposed inverse method offers a reliable and quick approach to determine the competitive adsorption isotherms for a specific SMB separation.

Relay simulated moving bed chromatography: concept and design criteria.

We present a new class of multicolumn chromatographic processes that change the classical way of handling the product outlets of simulated moving-bed (SMB) chromatography to avoid the use of flow controllers or an extra pump-the objective is to have just two- or three-way valves at a column outlet-while maintaining the analogy with the SMB in terms of displaced volumes of fluid per switch interval. In this class of processes the flow through a zone (or column) is always in one of the three states: (i) frozen, (ii) completely directed to the next zone (or column), or (iii) entirely diverted to a product line. We use the term relayed stream to refer to this particular type of manipulation of the outflow from a column. For this class of processes we derive a SMB analog-the R-SMB process-and demonstrate, under the framework of the equilibrium theory, that this process has the same separation region as the classical SMB for linear adsorption systems. In addition, the results from the equilibrium theory show that the R-SMB process consists of two distinct cycles that differ only in their intermediate sub-step: one cycle for selectivities α smaller than (3+√5)/2 and another cycle for larger values of α; in the former case no product stream is collected during the intermediate sub-step, whereas in the latter case both product streams are collected. We also examine the R-SMB process under conditions of finite column efficiency and compare its performance against those of the classical open- and closed-loop SMBs. Our simulation results show that the R-SMB process requires less desorbent and is more productive than the standard SMB processes under conditions of finite column efficiency and that the comparison increasingly favors the R-SMB as the column efficiency decreases.

Suppression of false arrhythmia alarms using ECG and pulsatile waveforms

In the context of the 2015 PhysioNet/CinC Challenge we present an algorithm to detect false arrhythmia alarms in the Intensive Care Unit(ICU). Our focus is on life threatening arrythmia alarms: asystole, extreme bradycardia, extreme tachycardia, ventricular tachycardia and ventricular fibrillation or flutter. Our method uses simultaneous ECG and pulsatile waveforms, photoplethysmogram or arterial blood pressure, to detect false alarms. QRS detectors produce for each signal a set of QRS detections that can be used to detect false alarms. Often the conclusions drawn from the results of QRS detectors on different signals are contradictory: some of the signals may be contaminated by noise or simply get lost for a while. Evaluating the signal quality of each waveform is necessary to can decide if we can trust the QRS detections obtained on that waveform. We describe a method to choose in each case which set of QRS detections should be used to conclude if the alarm is true or not. A set of rules is used for each alarm type.

Fetal beat detection in abdominal ECG recordings: global and time adaptive approaches.

We present a method for location of fetal QRS in maternal abdominal ECG recordings. This methods initial, global approach was proposed in the context of the 2013 PhysioNet/Computing in Cardiology Challenge where it was tested on the 447 four channel one-minute recordings.The first step is filtering to eliminate baseline wander and high frequency noise. Upon detection, maternal QRS is removed on each channel using a filter applied to the other three channels. Next we locate fetal QRS on each channel and select the channel with the best set of detections. The method was awarded the third-best score in the Challenge event 1 with 278.755 (beats/minute) and the fourth-best score on event 2 with 28.201 ms.The 5 min long recordings of the Abdominal and Direct Fetal ECG Database were used to further test the method. This database contains five recordings obtained from women in labor. Results in these longer recordings were not satisfactory. This appears to be particularly the case in recordings with a more clearly non-stationary nature. In a new approach to our method, some changes are introduced. Two features are updated over time: the filter used to eliminate maternal QRS and the channel used to detect fetal beats. These changes significantly improved the QRS detection performance on longer recordings, but the scores on the 1 minute Challenge recordings were degraded.

Single-Column Simulated Moving-Bed Process with Recycle Lag: Analysis and Applications

We report some conceptual and numerical results for a single-column chromatographic process analogous to a simulated moving bed (SMB) with asynchronous port switching. It is shown that the periodic state of the multi-column SMB unit is reproduced by a one-column chromatograph in which the part of its outlet stream that is not recovered as product is recycled back to the column with a lag of (N – 1)τ time units, where N and τ are, respectively, the number of columns and the switching interval of the equivalent SMB unit. This requirement holds true regardless of the port configuration or operational scheme of the analogous multi-column unit. The proposed one-column chromatographic process is a more compact, less expensive and simpler-to-operate alternative to the SMB. By changing the operation of the single-column chromatograph from the classical synchronous port switching to an optimized asynchronous Varicol mode, it is shown that one can easily compensate for a less perfect operation of the plug-flow device that implements the recycle lag.

MIXING ENHANCEMENT BY FREQUENCY-SELECTIVE CHAOTIC ADVECTION IN A 3-D TIME-PERIODIC STOKES FLOW

ABSTRACT A new 3-D periodic Stokes flow has been imagined and realized experimentally. It consists of axial Poiseuille flow superimposed on the 2-D tangential motion between two confocal ellipses that glide circumferentially so that the geometry is invariant. Chaotic streak lines obtained experimentally are compared to numerical simulations of this time-periodic flow. We next turn our attention to the problem of determining how to move the boundaries in order to obtain the most efficient mixing. Using a numerical experiment to study the advection of a passive scalar, we show that for a given 3-D mixer geometry and flow rate there is an optimum modulation frequency of the boundary displacement protocol for which the mixing process is most efficient. Furthermore, it is shown that chaotic advection can be regarded as a frequency-selective amplifier. This behavior is similar to that of fluid instability where external perturbations are amplified for a certain frequency range. For values above or below this range, perturbations are damped and the system is stable.

Detection of false arrhythmia alarms with emphasis on ventricular tachycardia

Our approach to detecting false arrhythmia alarms in the intensive care unit breaks down into several tasks. It involves beat detection on different signals: electrocardiogram, photoplethysmogram and arterial blood pressure. The quality of each channel has to be estimated in order to evaluate the reliability of obtained beat detections. The information about the heart rate from the different channels must be integrated in order to find a final conclusion. Some alarm types require particular detectors as is the case of ventricular fibrillation. To identify false ventricular tachycardia alarms we needed to classify heart beats as normal/ventricular. For that purpose we introduce a new feature, QRS polarity type. This feature was important in order to reduce misclassification of ventricular beats: there was an improvement in the ventricular tachycardia alarm true positive rate from 69% to 81%. However, the true negative rate was reduced from 95% to 69% and our global challenge score (real-time event) dropped from 79.02 to 74.28. Our challenge algorithm achieved the third best score in the 2015 PhysioNet/CinC challenge event 1 (real time).