Method for on-line prediction of conjunct polymer concentrations in hydrocarbon conversion processes

Abstract

A method for predicting conjunct polymer concentrations in spent ionic liquids during a continuous hydrocarbon conversion process is provided.

Description

Technical Field

[0001] The present disclosure generally relates to methods for analyzing spent ionic liquids in integrated catalytic hydrocarbon conversion systems, such as continuous alkylation systems. Background Art

[0002] Acidic ionic liquids can be used as catalysts in various hydrocarbon conversion reactions, such as the alkylation of isobutane with olefins, olefin oligomerization, paraffin disproportionation, paraffin isomerization, and aromatic alkylation. A byproduct of these reactions is the accumulation of conjunct polymers in the liquid catalyst over time. As will be appreciated, conjunct polymers are typically higher olefins, conjugated higher carbon cyclic hydrocarbons, formed as byproducts of various hydrocarbon conversion processes, including but not limited to alkylation, oligomerization, isomerization, and disproportionation.

[0003] Due to the olefin and diene functionality of the conjunct polymers, they have a strong affinity for acidic ionic liquid catalysts. This causes the catalyst to lose acidity as the amount of conjunct polymers in the ionic liquid catalyst increases. If the acidity of the ionic liquid catalyst decreases, the effectiveness of the catalyst in the reaction zone will also decrease.

[0004] The spent (or discarded) ionic liquid catalyst containing some conjunct polymers is typically recycled back to the reaction zone, and a side stream is typically diverted to the regeneration zone in order to maintain a constant level of catalyst activity.

[0005] There are several methods by which the ionic liquid catalyst can be regenerated. However, a determination must still be made as to whether the ionic liquid catalyst should be regenerated or whether the ionic liquid catalyst can be recycled back to the reaction zone.

[0006] The present disclosure relates to methods for determining analyte properties, such as the concentration of conjunct polymers in a waste ionic liquid, in real time or near real time using an infrared spectrometer integrated online with a hydrocarbon conversion system and coupled to an electronic controller that analyzes the information measured by the spectrometer. The infrared spectrum of a solution can be continuously monitored, and a chemometric model can be used to accurately and simultaneously characterize the quantitative chemical and / or physical properties of the analytes in the solution. Spectra can be obtained online from a flowing solution, allowing measurements to be made with little or no interruption to the hydrocarbon conversion process. Furthermore, the chemometric model can extract quantitative analyte information in real time or near real time, thereby enabling rapid feedback and control of process-related parameters and operations. Summary of the Invention

[0007] In one aspect, a method for predicting the concentration of conjunct polymers in a spent ionic liquid having an unknown concentration of conjunct polymers during a continuous hydrocarbon conversion process is provided, the method comprising: (a) separating an effluent from a reaction zone into a light fraction and a heavy fraction, the heavy fraction comprising the spent ionic liquid having an unknown concentration of conjunct polymers; (b) obtaining an infrared spectrum of each of a plurality of samples of the spent ionic liquid using an online infrared spectrometer configured with a measurement cell to allow the spent ionic liquid to flow therethrough; (c) separately determining the concentration of conjunct polymers in the spent ionic liquid by obtaining an infrared spectrum of each of the plurality of samples using an offline infrared spectrometer; (d) analyzing the infrared spectra obtained in (b) and (c) using multivariate chemometric techniques to provide a training data set; (e) generating a prediction model for the concentration of conjunct polymers based on the training data set; (f) applying the prediction model to the infrared spectra obtained in (b); and thereafter (g) quantitatively predicting the concentration of conjunct polymers in the spent ionic liquid during the continuous hydrocarbon conversion process. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A hydrocarbon conversion process according to some aspects of the disclosure herein is shown.

[0009] Figure 2 is a simplified schematic diagram of an exemplary configuration of a mixed polymer (CP) monitoring system according to aspects of the disclosure herein.

[0010] Figure 3 is a flow chart illustrating example steps that may be performed to analyze infrared spectroscopy information of spent ionic liquids according to some aspects of the present disclosure.

[0011] Figure 4 is a graph showing an overlay of Fourier transform infrared (FTIR) absorption spectra of conjunct polymers at various concentrations in spent ionic liquid.

[0012] Figure 5 is a graph showing the calibration of predicted conjunct polymer concentrations against actual conjunct polymers (CP) of spent ionic liquid samples using a partial least squares (PLS) chemometric model.

[0013] Figure 6 is a graph showing the validation of calculated conjunct polymer concentrations against actual conjunct polymer (CP) concentrations determined for offline spent ionic liquid samples.

[0014] Figure 7 is a graph showing a plurality of two-dimensional plots illustrating principal component analysis (PCA) scores obtained from a multivariate analysis of FTIR measurement results according to the present teachings.

[0015] Figure 8 is a graph showing the online prediction and offline determination of conjunct polymer concentration in the waste ionic liquid stream during alkylation over a two month period. DETAILED DESCRIPTION

[0016] definition

[0017] "Ionic liquid" refers to a salt (i.e., a composition comprising cations and anions) that is liquid at a temperature at or below about 150° C. (e.g., at or below about 120° C., 100° C., 80° C., 60° C., 40° C., or 25° C.). An ionic liquid is not considered a pure solution containing ions as solutes dissolved therein.

[0018] "Fresh ionic liquid" refers to an active ionic liquid catalyst that has not yet been used in any application, such as a new catalyst received from a supplier.

[0019] "Spent ionic liquid" refers to an ionic liquid catalyst that has been removed from a reaction zone, contains conjunct polymers, and has not been transferred to a regeneration zone, and may include regenerated ionic liquid catalyst that is reused as a catalyst in a reaction process.

[0020] "Regenerated ionic liquid" means an ionic liquid catalyst that has become spent and subsequently undergone a process to increase its activity level to a level higher than it was when the catalyst was spent. The activity of the regenerated ionic liquid is generally equal to or less than the activity of the fresh ionic liquid catalyst.

[0021] "Conjunct polymers" refers to materials containing olefins, conjugated hydrocarbons, and cyclic hydrocarbons that are formed as byproducts of various hydrocarbon conversion processes including, but not limited to, alkylation, oligomerization, isomerization, and disproportionation.

[0022] "In-line" means that the device (eg, an infrared spectrometer) is placed in a conventional fluid flow line.

[0023] "Offline" means that the device (eg, infrared spectrometer) is placed at an intermittent fluid flow, bypass flow, or sump (eg, the intermittent fluid is circulated in a loop) that is extracted from the regular fluid flow line.

[0024] "Continuous" means a system that operates without interruption or endlessly. For example, a continuous process for producing alkylate would be one in which the reactants are continuously introduced into one or more reactors and the alkylate product is continuously removed before the alkylation is stopped.

[0025] "Real-time" refers to measurements or processes that occur with a relatively short latency or recurrence. For example, a "real-time" measurement is one in which the total time interval between the start of measuring spectral information and the time when a parameter value or other quantity is calculated based on that information is one minute or less. Periodic real-time measurements are repetitive / periodic measurements in which the time interval between consecutive measurements is one minute or less.

[0026] A "near real-time" measurement is one in which the total time interval between the start of measuring the spectral information and the time when the parameter value or other quantity is calculated from that information is between 1 and 5 minutes. A periodic near real-time measurement is a repetitive / periodic measurement in which the time interval between consecutive measurements is between 1 and 5 minutes.

[0027] "Attenuated Total Reflectance" (ATR) is a sampling technique used in conjunction with infrared spectrometers that enables samples to be examined directly in the solid or liquid state without the need for further preparation.

[0028] A "distributed control system" (DCS) is a computerized control system for a process or plant in which autonomous controllers are distributed throughout the system, but there is a central operator monitoring and control.

[0029] A "zone" can refer to an area that includes one or more equipment items and / or one or more subzones. Equipment items can include one or more reactors or reactor vessels, heaters, exchangers, piping, pumps, compressors, and controllers. Additionally, an equipment item, such as a reactor, dryer, or vessel, can also include one or more zones or subzones.

[0030] introduction

[0031] According to the present disclosure, one or more of the following parameters can be controlled based on the predicted conjunct polymer concentration in the spent ionic liquid determined during the continuous hydrocarbon conversion process: the amount of spent ionic liquid returned to the reaction zone; the amount of spent ionic liquid delivered to the regeneration zone; the amount of fresh ionic liquid delivered to the reaction zone; and the amount of spent ionic liquid removed from the continuous hydrocarbon conversion process.

[0032] Predictions of conjunct polymer concentrations in spent ionic liquids according to the present disclosure are repeatable and can be repeated as long as the hydrocarbon conversion process continues.

[0033] It is contemplated that a desired conjunct polymer concentration range is maintained for the spent ionic liquid.

[0034] It is contemplated that a portion of the spent ionic liquid is passed to a regeneration zone to provide regenerated ionic liquid.

[0035] It is contemplated that if the conjunct polymer concentration in the spent ionic liquid is above a desired concentration range, the portion of spent ionic liquid returned to the reaction zone is reduced. If the conjunct polymer concentration in the spent ionic liquid is above a desired concentration range, fresh ionic liquid, regenerated ionic liquid, or both may also be delivered to the reaction zone.

[0036] Hydrocarbon conversion

[0037] The hydrocarbon conversion process involves contacting a hydrocarbon feed with an ionic liquid catalyst in a reaction zone under hydrocarbon conversion conditions. The effluent from the reaction zone is then separated into a heavy fraction containing spent ionic liquid catalyst and a light fraction containing reaction products. The separation can be performed by gravity, coalescence, or both, or by recovering droplets of spent ionic liquid catalyst in other ways.

[0038] Typical hydrocarbon conversion processes include alkylation, oligomerization, isomerization, and disproportionation.

[0039] Alkylation is commonly used to combine light olefins (e.g., mixtures of olefins such as propylene and butene) with isobutane to produce relatively high-octane branched-chain hydrocarbon fuels, including isoheptane and isooctane. Similarly, aromatic compounds (e.g., benzene) can be used in place of isobutane for alkylation reactions. When benzene is used, the products produced by the alkylation reaction are alkylbenzenes (e.g., toluene, xylene, ethylbenzene, etc.).

[0040] The process of oligomerizing light olefins (e.g., ethylene, propylene, and butenes) to produce high carbon number olefin products (e.g., C6+ olefins) is well known. Oligomerization processes have been used to produce high quality automotive fuel components and petrochemical products from ethylene, propylene, and butenes. Hydrocarbon feeds suitable for isomerization reactions include C2 to C23 olefins.

[0041] The isomerization of straight-chain alkanes into their branched isomers increases their octane rating, thereby increasing their value to the refinery. The isomerization process involves reacting one mole of a hydrocarbon (e.g., n-pentane) to form one mole of an isomer of that particular hydrocarbon (e.g., isopentane). The total number of moles remains the same throughout the process, and the products have the same carbon number as the reactants. Hydrocarbon feeds suitable for isomerization reactions include C3 to C23 paraffins.

[0042] The disproportionation of paraffins (e.g., isopentane) involves reacting two moles of the hydrocarbon to form one mole each of two different products, one product having a higher carbon number than the starting material and the other having a lower carbon number than the starting material. The total number of moles in the system remains the same throughout the process, but the carbon number of the products differs from that of the reactants. Hydrocarbon feeds suitable for the disproportionation reaction include C2 to C23 paraffins. Feeds containing two or more paraffins are also acceptable.

[0043] The hydrocarbon conversion conditions depend on the specific hydrocarbon conversion process. The reaction temperature is generally in the range of -20°C to 250°C. The pressure is generally in the range of 0 MPa(g) to 13.8 MPa(g).

[0044] Figure 1 One aspect of the hydrocarbon conversion process of the present disclosure is shown, illustrating an example of an ionic liquid catalyzed alkylation zone for producing alkylate gasoline blending components. Figure 1 , a hydrocarbon feed 10 is passed to a reaction zone 12. The hydrocarbon feed 10 typically comprises a mixture of olefins, paraffins, and isoparaffins, and these can be added separately at one or more locations in the reaction zone 12. Thus, the hydrocarbon feed 10 comprises an olefin stream of olefins. In addition, an isoparaffin stream 14 of isoparaffins is also passed to the reaction zone 12. An ionic liquid stream 16 is also passed to the reaction zone 12. The reaction zone 12 comprises at least one reactor for the alkylation reaction.

[0045] In general, the alkylation process includes transferring isoparaffins and olefins to an alkylation reactor, wherein the alkylation zone 12 includes an ionic liquid catalyst to react the olefins with the isoparaffins to form alkylates. The paraffins used in the alkylation process preferably include isoparaffins having 4 to 10 carbon atoms (e.g., 4 to 8 carbon atoms, or 4 to 5 carbon atoms). The olefins used in the alkylation process preferably have 2 to 10 carbon atoms (e.g., 3 to 8 carbon atoms, or 3 to 5 carbon atoms). The isoparaffins have 4 to 10 carbon atoms, and the olefins have 2 to 10 carbon atoms. According to one or more aspects of the present disclosure, the alkylation process elevates relatively low-value C4 carbons to higher-value alkylates. In this regard, a particular aspect is the alkylation of butanes with butenes to form C8 compounds. Preferred products include trimethylpentane (TMP), and although other C8 isomers are produced, a competitive isomer is dimethylhexane (DMH). The quality of the product stream can be measured by the ratio of TMP to DMH, with a high ratio being desirable.

[0046] Typical alkylation reaction conditions include temperatures in the range of -20°C to the decomposition temperature of the ionic liquid or -20°C to 100°C (e.g., -20°C to 80°C, or 0°C to 80°C, or 20°C to 80°C, or 20°C to 70°C, or 20°C to 50°C). Ionic liquids can also solidify at moderately low temperatures, so it is preferred to have ionic liquids that remain liquid over a reasonable temperature range. In some aspects, cooling may be required. If cooling is required, any known method can be used to provide cooling. The pressure is typically in the range of 0.1 to 8.0 MPa(g) or 0.3 to 2.5 MPa(g). The pressure is preferably sufficient to maintain the reactants in the liquid phase. The residence time of the reactants in the reaction zone 12 is in the range of seconds to hours (e.g., 0.5 minutes to 60 minutes, or 1 minute to 60 minutes, or 3 minutes to 60 minutes).

[0047] Due to the low solubility of hydrocarbons in ionic liquids, olefin-isoparaffin alkylation, like most reactions in ionic liquids, is typically biphasic and occurs at the interface in the liquid state. As is common in aliphatic alkylations, catalytic alkylation reactions are typically carried out in the liquid hydrocarbon phase in a batch, semi-batch, or continuous system using a single reaction stage. The isoparaffin and olefin can be introduced separately or as a mixture. The molar ratio of isoparaffin to olefin is in the range of 1:1 to 100:1 (e.g., 2:1 to 50:1, or 2:1 to 20:1).

[0048] Ionic liquids contain organic cations and anions. Suitable organic cations include nitrogen-containing cations and phosphorus-containing cations. Suitable organic cations include ammonium cations, pyridinium cations, imidazolium cations, and phosphonium cations.

[0049] Suitable anions include metal halide anions, non-metal halide anions, and combinations thereof. The metal halide anions and / or non-metal halide may include at least one halide selected from F, Cl, Br, and I. In some aspects, the metal halide anions include metal chlorides. In some aspects, the non-metal halide anions include non-metal fluorides. Exemplary non-metal halide anions include tetrafluoroborate, hexafluorophosphate, and bis(trifluoromethanesulfonimide).

[0050] The metal in the metal halide anion can include a metal selected from Group 13 metals, transition metals, or combinations thereof. In some aspects, the metal can be selected from aluminum, gallium, indium, titanium, zirconium, chromium, iron, copper, zinc, tin, and combinations thereof. In some aspects, the metal halide anion includes aluminum halide.

[0051] In some aspects, the metal halide anion can be selected from chloroaluminates, chlorogallates, chloroindiumates, chlorotitanates, chlorozirconates, chlorochromates, chloroferrates, chlorocuprates, chlorozincates, chlorostannates, and combinations thereof. In some aspects, the metal halide anion can include chloroaluminates. In some aspects, the metal halide anion can include [Al2Cl7] - 、[AlCl4] - or [Ga2Cll7] - .

[0052] In some aspects, the ionic liquid is selected from the group consisting of tetraalkylammonium chloroaluminates, 1-alkylpyridinium chloroaluminates, 1,3-dialkylimidazolium chloroaluminates, tetraalkylphosphonium chloroaluminates, and combinations thereof. Exemplary ionic liquids include 1-butylpyridinium chloroaluminate, 1-butyl-3-methylimidazolium chloroaluminate, and combinations thereof.

[0053] return Figure 1 The effluent 20 (including hydrocarbon conversion products, ionic liquid catalyst, and unconverted reactants, any co-catalysts such as hydrogen chloride, organic chloride, or other compounds) is passed from the reactor 12 to a separation zone 22 having one or more separation vessels. In the separation zone 22, the effluent 20 separates into a hydrocarbon phase and an ionic liquid phase. This separation can be a phase separation due to a density difference between the hydrocarbon phase and the ionic liquid phase, but other methods, including, for example, coalescing materials, can also be used. Thus, the alkylation effluent stream 26 is sent to a product recovery section 28. A spent ionic liquid catalyst stream 30, which typically includes a certain amount of conjunct polymers, is also recovered from the separation zone 22. A first portion 32 of the spent ionic liquid catalyst stream 30 can be directly recycled to the reaction zone 12, while a second portion 34 of the spent ionic liquid catalyst stream 30 is passed to a regeneration zone 36 to remove at least some of the conjunct polymers from the ionic liquid catalyst. The regenerated ionic liquid 38 can be passed back to the reaction zone 12.

[0054] In product recovery zone 28, alkylation effluent stream 26 is separated into alkylation product 42 and recycle hydrocarbons 40, which include unreacted hydrocarbons (including iC4) and some lower carbon compounds (including HCl), which are then returned to reaction zone 12 as a recycle stream.

[0055] The conjunct polymers are bound to the spent ionic liquid catalyst as an integral compound. Simple hydrocarbon solvent extraction does not wash the conjunct polymers from the spent ionic liquid catalyst. It is believed that the most efficient and effective method for reducing the conjunct polymer content in spent ionic liquid catalyst is to convert the conjunct polymer material into extractable hydrocarbons (i.e., light hydrocarbon gases or saturated hydrocarbons that are not readily soluble in ionic liquids) and subsequently separate or extract the hydrocarbons from the regenerated ionic liquid catalyst. During the regeneration process, a portion of the conjunct polymers can be hydrocracked into lower carbon species (C1-C4 hydrocarbons) that can be incorporated into the exhaust gas, a portion of the conjunct polymers can be hydrocracked into a liquid hydrocarbon stream that can be incorporated into the alkylate gasoline product, and / or a portion of the conjunct polymers can be hydrocracked into HCl gas.

[0056] As long as the conjunct polymer content in the spent ionic liquid is low enough that the effectiveness of the ionic liquid is not significantly negatively affected, the spent ionic liquid can be returned to reaction zone 12 and reused as an ionic liquid catalyst in reaction zone 12. The present disclosure relates to the use of an on-line infrared spectrometer to monitor the amount of conjunct polymers in the spent ionic liquid in real time or near real time.

[0057] The infrared spectrometer is placed in-line, meaning that the spectrometer is placed (i.e., located) in the line through which the waste ionic liquid flows or in the separation zone 22, where the waste ionic liquid is a separate phase, at a location where spectral information can be directly measured. Thus, the spectrometer can be placed in any of the lines 30, 32, and 34 used to return the waste ionic liquid to the reaction zone 12. It is contemplated that more than one infrared spectrometer may be used, and that the spectrometers may be located at various locations throughout the process.

[0058] For example, when the current process indicates that the conjunct polymer concentration is high, a first portion of the spent ionic liquid catalyst can be returned to reaction zone 12 while a second portion of the spent ionic liquid catalyst is passed to regeneration zone 36. Once the concentration of conjunct polymers returns to a desired level, the flow of spent ionic liquid catalyst can be adjusted accordingly, such that more of the spent ionic liquid catalyst is recycled back to reaction zone 12 and less of the spent ionic liquid catalyst is passed to regeneration zone 36, or all of the spent ionic liquid catalyst is returned to reaction zone 12.

[0059] Figure 2is a simplified schematic diagram of a configuration of a system for monitoring mixed polymers according to some aspects of the present disclosure herein. Waste ionic liquid can be directed to an offline sampling station and / or an online infrared spectrometer. The spectrometer can be an FTIR spectrometer that receives reflected radiation and generates spectral information of the reflected radiation. A suitable FTIR spectrometer for use as a detector is a Bruker ALPHAII FTIR (available from Bruker Optics of Billerica, Massachusetts) with a deuterated triglycine sulfate (DTGS) infrared sensor, but many other FTIR spectrometers can also be used. The spectrometer is connected to a personal computer (PC) via a data link, such as an Ethernet cable or a Wi-Fi network. The PC can be located in an analyzer building separate from the location of the spectrometer. The PC can be connected to a distributed control system (DCS) via a data link.

[0060] Infrared spectroscopy and measurement systems

[0061] As described above, conjunct polymers in waste ionic liquids can be quantitatively measured using an online infrared spectrometer. Preferably, an internal reflection FTIR method is used to perform an "in situ" measurement of the infrared spectrum absorbed by the waste ionic liquid in or leaving one or more processing zones or containers. The internal reflection FTIR method allows for in situ measurement of the infrared spectrum absorbed by the reaction solution by placing a sensor probe in, on, or near the processing line or processing container so that it is immersed in the waste ionic liquid; or placing the sensor probe in a direct or reflected line of sight of the waste ionic liquid, thereby allowing for direct scanning of the waste ionic liquid in essentially real time without removing a sample of the solution from the container or processing line containing the solution. Advantageously, the in situ measurement provides real-time or near real-time measurement of the waste ionic liquid.

[0062] Generally speaking, internal reflection involves the process of modulating an infrared beam using an interferometer. The modulated beam reflects off a sample and returns to a detector, where the absorbed spectral regions and the absorption intensity within those regions are determined. One technique for implementing internal reflection methods is attenuated total reflectance (ATR) spectroscopy, which measures absorbance in a thin layer of the sample in contact with the sampling surface of a sensor device. Specifically, a sensor probe is placed in direct contact with the sample. A modulated infrared beam is transmitted from an FTIR spectrometer to a sensor probe, where it transmits through the sampling surface on the probe, penetrating into the thin layer of the sample in contact with the probe's sampling surface and reflecting back into the sensor probe. Significantly, the beam is modified by the sample due to its absorption properties. This modified beam is then optically transmitted to the detector of the FTIR spectrometer. Depending on the chosen ATR probe (i.e., the optical properties and geometry of the sampling surface), the modulated infrared beam may reflect off the sample layer and sampling surface multiple times before finally returning to the sensor probe, providing additional data to the detector. Therefore, ATR probes are often described by the number of reflections occurring within the sample layer. Preferably, the ATR probe utilizes at least about 3, more preferably at least about 6, and more preferably at least about 9 reflections or more.

[0063] Preferably, the sampling surface of the ATR probe is composed of diamond. ATR probes comprising a diamond sampling surface may also include an additional optical element that serves as a support for the diamond and is used to transmit and focus the modulated infrared light beam to and from the diamond sampling surface. Because the second optical element does not come into contact with the reaction solution, it is less important that the second optical element be as corrosion-resistant and abrasion-resistant as the sampling surface. Zinc selenide crystals have optical qualities similar to diamond, but at a significantly lower cost. Therefore, zinc selenide can be used as the additional optical element.

[0064] The sampling surface of the ATR probe can be concave, convex, or have a relatively flat surface curvature. Preferably, the sampling surface of the ATR probe is relatively flat. Without being bound by a particular theory, it is believed that a sampling surface with a significant curvature tends to encourage particles to adhere to the sampling surface, thereby interfering with the sensor.

[0065] FTIR spectrometers detect the intensity or amplitude of a modified light beam in the infrared region and convert the data into an absorption spectrum, i.e., absorbance versus wavenumber. FTIR spectrometers typically use two types of detectors: mercury cadmium telluride (MCT) detectors or deuterated triglycine sulfate (DTGS) detectors. While MCT detectors tend to be faster and more sensitive than DTGS detectors, they typically require liquid nitrogen or thermoelectric cooling systems, making them inconvenient and economically undesirable.

[0066] Typically in the spectral region of 2 to 50 microns (i.e., 200 cm -1 Up to 5000cm -1 or 650cm -1 Up to 4000cm -1 The waste ionic liquid is sampled at a wavenumber of 2, 4, 8 or 16 (i.e., sample data is collected over a discrete range of 2, 4, 8 or 16 wavenumbers, where the resolution is inversely proportional to the wavenumber resolution). Infrared spectroscopy is a continuous spectrum, but for analytical reasons, discrete wavenumbers or groups of wavenumbers are typically measured. The wavenumber resolution (i.e., the range of wavenumbers that are grouped together for each discrete measurement) can be increased or decreased to affect the signal-to-noise ratio of the FTIR spectrometer. That is, as the value of the wavenumber resolution decreases, more measurements are taken of the spectrum, and the resolution of the spectrum increases. However, an increase in the wavenumber resolution typically also results in a corresponding increase in the level of "noise." In general, FTIR spectroscopy uses a wavenumber resolution of 2, 4, 8 or 16 (i.e., sample data is collected over a discrete range of 2, 4, 8 or 16 wavenumbers, where the resolution is inversely proportional to the wavenumber resolution). Typically, a wavenumber resolution of 4 appears to provide a spectrum with fairly good resolution while minimizing the amount of "noise." However, the wavenumber resolution can be varied without departing from the scope of the present disclosure.

[0067] In addition, FTIR spectroscopy typically utilizes multiple scans to provide additional interferometric data, i.e., intensity versus wavenumber data used in the Fourier transform to generate spectral data (i.e., absorbance versus wavenumber). If the number of scans is set to 32, for example, the spectrometer will scan the entire specified wavelength range 32 times and generate 32 interferograms, or 32 intensity measurements per wavenumber or, more precisely, per wavenumber grouping, as determined by the wavenumber resolution. The Fourier transform then combines the intensity data and converts the 32 interferograms into a single absorption spectrum. The number of spectra, i.e., the number of scans, may also affect the signal-to-noise ratio. Generally, approximately 32 scans can be sampled, with a new spectral measurement generated approximately every 25 seconds.

[0068] Chemometric-based analysis of infrared spectral information

[0069] A single analyte will produce a spectrum with an absorbance curve characteristic of that analyte. In other words, the spectrum contains absorbance features that may be associated with the analyte. Therefore, a mathematical model representing the relationship between the concentration of the analyte and the absorbance curve can be used to determine the concentration of the analyte. The mathematical model can be formed by measuring the spectra of many standard samples with known concentrations and using a number of correlation methods to mathematically correlate the concentration with the absorbance curve. Unfortunately, the characteristic spectra of analyte mixtures, such as waste ionic liquids containing mixed polymers, are more complex because the characteristic absorption spectra of the various analytes are significantly broad and overlap. This overlap hinders the use of simple univariate correlation methods to quantify the analytes in the reaction mixture. This problem can be overcome by applying more powerful multivariate mathematical correlation techniques to analyze spectral data. These multivariate mathematical techniques are collectively referred to as chemometrics when applied to chemical analysis. This technique uses complex mathematical algorithms (such as matrix-vector algebra and statistics) to extract quantitative information (e.g., concentration) from highly convoluted or statistically obfuscated data (e.g., spectra obtained from analyte mixtures) to form a mathematical model, which is also referred to as a chemometric model that represents quantitative information as it varies with the spectrum. Many multivariate mathematical techniques have been developed, such as K-nearest neighbor analysis (KNN), hierarchical cluster analysis (HCA), principal component analysis (PCA), partial least squares (PLS) analysis, and principal component regression (PCR) analysis. Commercially available software packages are capable of performing many of the above multivariate mathematical related techniques.

[0070] Commercially available FTIR spectrometers typically include chemometric analysis software. Specifically, PLS and PCR are commonly used to determine a chemometric model and apply it to FTIR spectral measurements to calculate the properties of the measured sample. Of the two, PLS is most commonly applied to FTIR spectral data because it generally provides the most accurate chemometric model. PLS allows each analyte to be modeled individually, requiring only knowledge of the specific analyte being modeled. In other words, it is not necessary to know the concentration of each absorbing analyte, as long as each absorbing analyte is represented in the standard used to develop the chemometric model. Advantageously, the standards can be obtained directly from the process and do not need to be prepared separately, allowing the impurity profile of the waste ionic liquid to be considered when determining the chemometric model for each analyte to be measured. However, it should be noted that the absorbance of a spectral region is generally nonlinear with respect to concentration. Therefore, a chemometric model that relates concentration to absorption spectrum should be developed for each analyte in the reaction solution within a specific concentration range. In other words, the standards used in chemometric analysis should represent a matrix of the concentrations of each analyte in the reaction solution.

[0071] Therefore, generally, an infrared spectrometer is used to analyze a number of standards to measure the spectrum of each standard. The concentration of a specific analyte can then be mathematically modeled based on the obtained spectrum, that is, an algorithm is formed to relate the concentration to the spectrum. Although any multivariate mathematical calibration technique can be used, a preferred embodiment uses a PLS method to model the spectrum based on concentration. The number of standards used is preferably at least about 10, more preferably at least about 20. Generally, the accuracy of the model increases as the number of standards used to generate the model increases. Therefore, the number of standards used to generate the model can be as high as 50 or more. Such standards can be prepared mixtures, or they can be samples of the specific process mixture to be analyzed. However, as mentioned above, it is preferred to use a process mixture so that the impurity profile is taken into account in the PLS analysis when generating the chemometric model. Standard analytical techniques such as high performance liquid chromatography (HPLC) can be used to measure the concentration of the modeled analyte in each standard offline. Therefore, a chemometric model can be generated using partial least squares regression analysis of the spectrum obtained from the reaction mixture based on online spectral measurements and offline HPLC concentration measurements based on a batch or continuous alkylation process.

[0072] As mentioned above, FTIR is at the wavelength corresponding to 200 cm -1 Up to 5000cm -1 and more preferably 650 cm -1 Up to 4000cm -1 The waste ionic liquid is scanned over a spectral range of wavelengths of 1000 nm to 1000 nm. Although the entire spectral region scanned can be used in the PLS analysis, in general, when modeling mixed polymer analytes, the spectral region of mixed polymers considered in the PLS analysis is preferably 800 cm -1 to 1800cm -1 (For example, 1300cm -1 to 1400cm -1 ).

[0073] Thus, using PLS analysis techniques, chemometric models can be developed for determining the concentration of mixed polymer analytes from absorbance spectra, and the chemometric models can be used in conjunction with FTIR spectroscopy to provide real-time concentration data for process mixtures from batch or continuous processes, thereby allowing for improved studies of reaction kinetics, improved reaction control, and, in the case of batch processes, more accurate and timely determination of reaction endpoints.

[0074] For example, using the above-described techniques, a chemometric model was developed using FTIR spectroscopy and a diamond composite ATR probe, allowing the measurement of conjunct polymer concentrations in spent ionic liquids over a concentration range from about the detection limit (currently about 50 ppm) to about 7%, with an average PLS error of less than about 0.2% for the continuous alkylation process.

[0075] Figure 3 is a flow chart illustrating example steps that may be performed to analyze infrared spectroscopy information of spent ionic liquids according to some aspects of the present disclosure.

[0076] Example

[0077] The following illustrative examples are intended to be non-limiting.

[0078] Example

[0079] In a pilot plant alkylation unit using acidic ionic liquids to produce alkylate gasoline blending components, a Bruker FTIR spectrometer ALPHA II equipped with an ATR cell was installed in the waste ionic liquid stream. Online FTIR spectral data were collected continuously at one-minute intervals using a flow cell with a DTGS detector. For each spectrum, the 4 cm -1 32 scans were collected at 400 nm resolution.

[0080] In addition, 360 spent ionic liquid samples were collected over a two-month period and analyzed on a Bruker ALPHAIIATR-FTIR spectrometer at 4 cm -1 The samples were analyzed offline using 8 scans at a resolution of 100 Å to obtain the conjunct polymer concentrations in the spent ionic liquid (the conjunct polymer concentration range was 0 to 6 wt. %). Figure 4 is a graph showing the overlay of FTIR absorption spectra of mixed polymer calibration samples.

[0081] By utilizing multivariate mathematical analysis of spectral data from a continuous alkylation process at 800 cm -1 to 1800cm -1 FTIR spectral data were obtained in a spectral region including the absorption bands associated with the mixed polymer components (1300 cm -1 to 1400cm -1 ) The cross-validated PLS regression model for quantifying conjunct polymers in waste ionic liquids was developed using Bruker QUANT2 chemometrics software (v. 7.3). A set of 178 samples representing a variation in conjunct polymer composition was created and used to develop a cross-validation calibration and as a performance test. Another set of 169 samples was not included in the calibration and was validated within the calibration range.

[0082] Figure 5 is a graph showing the calibration of predicted conjunct polymer concentrations against actual conjunct polymers of spent ionic liquid samples using the PLS chemometric model. Figure 5 The correlation coefficient (R 2 ) is 98.92 and the root mean square error of estimation (RMSEE) is 0.111, which indicates promising findings.

[0083] Figure 6 is a graph showing the validation of calculated conjunct polymer concentrations against actual conjunct polymer concentrations determined for spent ionic liquid samples. Figure 5 The correlation coefficient (R 2 ) is 98.81 and the root mean square error of prediction (RMSEP) is 0.108. The validation model demonstrates that the conjunct polymer concentration in the alkylation stream can be measured online with very high accuracy.

[0084] To further investigate the concentration threshold for reliable testing, principal component analysis (PCA) was applied to two series of mixed polymer datasets. Figure 7 In Figure 1, the lower trace shows the first background data set, while the upper trace shows the second background data set after clearing the ATR window of the FTIR spectrometer and then collecting a new background data set. PCA analysis of the first derivative IR with different analyses showed that these two data sets did not affect the PLS model. However, once the final PLS model was established, PCA analysis was used to calculate and predict the unknown online mixed polymer sample based on the IR absorption spectrum.

[0085] Figure 8 is a graph showing the concentration of conjunct polymers in a spent ionic liquid stream during alkylation over a two-month period. As shown, the trend of the predicted online conjunct polymer concentration values closely correlates with the offline conjunct polymer concentration values. The presence of isobutane and the buildup of contaminants on the ATR window of the FTIR spectrometer did not appear to affect the analysis, even with varying infrared backgrounds. The results demonstrate that the PLS model is able to quantify sample components as the correlation between concentration and absorbance changes, and even when the background changes. Components may cause peak shifts or changes in the mixture spectrum, but the spectral changes present in the standard accurately represent the expected changes in the unknown conjunct polymer sample. The results also demonstrate that PLS multivariate calibration provides highly accurate conjunct polymer measurements and allows operators and engineers to respond more quickly to even small changes in ionic liquid catalyst composition and / or activity.

Claims

1. A method for predicting the concentration of conjunct polymers in a spent ionic liquid having an unknown concentration of conjunct polymers during a continuous hydrocarbon conversion process, the method comprising: (a) contacting a hydrocarbon feed with an ionic liquid in a hydrocarbon conversion reaction zone under hydrocarbon conversion conditions to form a mixture comprising reaction products and spent ionic liquid containing conjunct polymers; (b) separating the mixture into a light fraction and a heavy fraction, wherein the heavy fraction comprises a waste ionic liquid having an unknown concentration of conjunct polymers, (c) obtaining an infrared spectrum of each of the plurality of samples of the waste ionic liquid from (b) using an online infrared spectrometer, wherein the online infrared spectrometer is configured with a measurement flow cell to allow the waste ionic liquid to flow therethrough, wherein the infrared spectrum at 800 cm -1 to 1800cm -1 The absorbance value indicating the mixed polymer is measured in each infrared spectrum within the wavenumber range; (d) collecting a sample of the spent ionic liquid from (b) at each of a plurality of corresponding time points during the continuous hydrocarbon conversion process, and determining the conjunct polymer concentration in the spent ionic liquid offline by obtaining a corresponding infrared spectrum for each sample using an offline infrared spectrometer, wherein the concentration of the conjunct polymers in the spent ionic liquid is determined by measuring the infrared spectrum at 800 cm -1 to 1800cm -1 The absorbance value indicating the mixed polymer is measured in each infrared spectrum within the wavenumber range; (e) analyzing the infrared spectra acquired in (c) and (d) using a multivariate chemometric technique to provide a training data set, wherein the multivariate chemometric technique includes partial least squares (PLS) analysis; (f) generating a prediction model for mixed polymer concentration based on the training data set, wherein the prediction model is built using principal component analysis (PCA); (g) applying the prediction model to the infrared spectrum obtained in (c); and thereafter (h) quantitatively predicting the conjunct polymer concentration in the spent ionic liquid during the continuous hydrocarbon conversion process.

2. The method of claim 1, wherein the hydrocarbon conversion process comprises at least one of alkylation, disproportionation, isomerization, and oligomerization.

3. The process of claim 2, wherein the alkylating comprises contacting an isoparaffin feed having from 4 to 10 carbon atoms with an olefin feed having from 2 to 10 carbon atoms in an alkylation zone in the presence of an ionic liquid under alkylating conditions to produce an alkylate.

4. The method of claim 1, wherein acquiring the infrared spectrum in (c) and / or (d) comprises acquiring an attenuated total reflectance Fourier transform infrared spectrum.

5. The method of claim 1, further comprising: Based on the predicted conjunct polymer concentration in (h), at least one of the following is controlled: (i) the amount of waste ionic liquid returned to the reaction zone; (ii) the amount of spent ionic liquid delivered to the regeneration zone; (iii) the amount of regenerated ionic liquid returned to the reaction zone; (iv) the amount of fresh ionic liquid delivered to the reaction zone; and (v) the amount of spent ionic liquid catalyst removed from the continuous hydrocarbon conversion process.

6. The method of claim 1, further comprising: If the conjunct polymer concentration in the waste ionic liquid is above a desired range, the amount of waste ionic liquid returned to the reaction zone is reduced.

7. The method of claim 1, further comprising: If the conjunct polymer concentration in the spent ionic liquid is above the desired range, fresh ionic liquid, regenerated ionic liquid, or both are delivered to the reaction zone.

8. The method of claim 1, wherein predicting the conjunct polymer concentration in the spent ionic liquid during the continuous hydrocarbon conversion process is performed in real time or near real time.

9. The method of claim 1, wherein in (d), at least 50 samples are collected.

10. The method of claim 1, wherein the spent ionic liquid containing conjunct polymers has a conjunct polymer concentration of 50 ppm to 7% by weight of the conjunct polymers.

Images