Designs and methods for interfacing SFC with MS

Designs and methods for interfacing SFC with MS

1

2

Abhijit Tarafder

3

∗

Waters Corporation, Milford MA 01757, USA

Abstract

4

5

Hyphenating SFC with MS is now routinely performed in analytical laboratories.

6

Major instrument providers supply commercial solutions for coupling SFC and MS,

7

which has facilitated wider adoption of the technology. The current status, however,

8

could be achieved based on the work done by many researchers over decades. Interfac-

9

ing SFC with MS posed some unique challenges, compared to interfacing MS with LC

10

or GC, demanding special solutions. Several interface designs were tried and tested

11

over the years before suitable solutions could be detected. Additional measures, such

12

as (a) mixing SFC mobile-phase with an additional liquid solvent at the column outlet,

13

and (b) heating the interfacing device, had to be adopted to address some specific

14

challenges. Although such modifications and measures look diverse, there is one factor

15

that drove most of them - compressibility of SFC mobile-phase. There are two objec-

16

tives of this review - (1) to compile various insights which were reported on describing

17

and optimizing SFC-MS interfacing processes, and (2) to link these insights with the

18

fundamental issue of solvent compressibility. ∗

Corresponding author E-mail:abhijit [email protected]

1

19 20

Keywords: Supercritical fluid chromatography, mass spectrometry, SFC-MS, interfacing, coupling, mobile-phase, solvent, compressible, expansion

2

21

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

1

Introduction

Interest in SFC-MS has steadily increased over the last few years supported by commercial availability of more robust and reliable analytical SFC systems. Interfacing SFC with MS does not require special technical skills anymore as the solutions and guidance from major instrument providers sufficiently address most situations. During hyphenation, apart from the required SFC and MS-related method parameters, the only interface-specific parameters that need to be supplied are - the flowrate and the composition of a ”make-up” solvent. Make-up solvent is the liquid organic solvent(s) added to SFC mobile-phase at the column exit, to improve MS detection. There is also a practice of heating the interface connector employing thermal sleeve(s). Empirical guidelines are available on selecting the make-up flowrates and compositions, and also on selecting connector heater temperatures, which are helpful most of the times. Sometimes, however, the situation can be baffling if the basic mechanisms that control the interfacing process are not properly understood. The purpose of this review is to provide some physical explanations behind the empirical guidelines. Objectively, the only task of an interfacing device between chromatography and MS is to transfer the analyte molecules eluting from the column to the MS inlet, while ensuring that the separation achieved through chromatography is maintained. Such a device can be a simple connecting tube - like in GC-MS and LC-MS. For SFC-MS, however, the situation is not so simple. SFC requires to be conducted under highly pressurized conditions and because MS is a low-pressure detector, analytes eluting from SFC column must be depressurized before being supplied to the MS inlet. This depressurization must be done in a controlled way, otherwise it can destroy the resolution achieved by chromatography and also cause poor MS signals. This is the most complex problem related to SFC-MS interfacing and it exists because of the highly compressible nature of SFC mobile-phase. The other manifestation of the challenge of working with a compressible solvent is the requirement of employing an automated back-pressure regulator (ABPR), which is a control valve employed to maintain a set pressure inside the system while allowing the system to work with any flowrates. ABPRs are employed after the column for understandable reason and may come with voluminous designs that can add dispersion to analyte bands. To address this issue sometimes a part of the mobile-phase is diverted to the MS by splitting the main stream. Most challenges related to SFC-MS interface design can be categorized under - (a) how to design a pressurized system where the pressure controlling device is not contributing significantly to system dispersion, and (b) how to manage solvent decompression properly so that chromatographic fidelity is not compromised. Solutions adopted to address SFC3

60

MS interfacing problems inherently address either one or both the issues. Based on this insight the rest of the review is divided in two major parts, describing - (a) technological solutions developed to reduce the effect of additional system volume in form of ABPR (in section 2), and (b) methods and designs developed to execute solvent decompression without losing chromatographic resolution (in section 3). Note that both components are critical to successfully couple SFC with MS.

61

2

55 56 57 58 59

SFC-MS coupling strategies

68

One of the main issues that influenced strategies to fluidically couple SFC with MS was of maintaining a set pressure inside SFC system while not adding a significant volume by placing a bulky pressure-controlling device before MS. Excellent reviews on this topic has been published before [1, 2]. The design approaches taken to accomplish this task can be divided in two broad categories - (a) full-flow introduction, and (b) split-flow introduction. Schematic diagrams representing the two approaches are presented in Figure 1 and a review from different reports that applied these designs are presented in the rest of this section.

69

2.1

62 63 64 65 66 67

70 71 72 73 74 75 76 77 78 79

Full-flow introduction

Directing the full flow from SFC to MS was the earliest design adaptation for SFC-MS interfacing, reported by Randall and Wahrhaftig [3]. Referring the design as dense gaschromatography/mass-spectrometer interface, the authors expanded the dense gas in a nozzleskimmer-collimator system. The advantage of full-flow is the ability to introduce all or most of the analyte molecules for detection - potentially increasing sensitivity for mass-flow sensitive ionization methods e.g. APCI (atmospheric pressure chemical ionization). The challenge, however, is to design an interface which is low-volume, robust and does not put any constraint on SFC method design. Based on the reported design alternatives, full-flow is achieved through - (1) employing a capillary restrictor, (2) employing a liquid pump to control SFC system pressure, and (c) employing low-volume ABPR.

4

80

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113

2.1.1

Full-flow through capillary restrictor

Directing full-flow through a capillary restrictor, which acts as a passive pressure regulator, was first introduced by Smith et. al. [4] in 1982. The process was called direct fluid introduction (DFI). The authors reported coupling an open-tubular capillary SFC with a chemical-ionization MS by employing a short restrictor with a carefully designed pinchedend. Smith and co-workers reported a series of pioneering work in SFC-MS interfacing [4–6]. The group designed and tested various restrictor geometries to optimize MS performance [6]. More detailed discussion on these designs is presented in section 3.2. Concurrently, Arpino and co-workers reported on the basic mechanism of SFC-MS interfacing and ionization processes [7–11]. Figures 1(a) and (b) demonstrate the interfacing scheme of full-flow restrictors. Heating the tip of the restrictor, as schematically shown in Figure 1(a), or sometimes heating the entire restrictor, was practiced to counter SFC mobile-phase cooling due to rapid decompression [8, 12, 13]. The heating process also arrested the analytes from precipitating. Apart from heating, multi-step expansion of mobile-phase to address analyte precipitation, was also reported. E.g. Cousin et al [8] used an expansion chamber before introducing the eluting fluid to a heated restrictor. Morgan et al [14] reported employing a variable restrictor, which was placed after a photo-diode array detector inline with the MS. The main advantage of using a restrictor is its relatively simple design. The main challenges, however, are reduced operational flexibility and lack of robustness depending on restrictor design [6]. Pressure drop through a passive regulator is controlled by the solvent properties - density and viscosity, and also the mass flow rate. As a result, whenever the mobile-phase components or the composition or the flowrate is changed, pressure inside the SFC system will change. In other words, controlling the pressure inside SFC system, independent of flowrate and/or composition, becomes impossible with a fixed restrictor. There are, however, conditions where this limitation is not a major problem or can be addressed through other innovative ways. For example, Pinkston noted [2] that such devices can still be useful if operated with low compressibility SFC mobile-phases, which is more ”liquid-like”, so that pressure variation will not have a strong effect on the separation. One such example was reported by Hoke et al [15] who employed a mobile-phase with 35/65 (v/v,%) CO2 /methanol composition. An innovative way to employ fixed restrictor but still able to maintain a set column outlet pressure under varying co-solvent compositions was presented by Hayward and Han [16]. The SFC column effluent was pre-heated before sending to the restrictor (Figure 1(b)) which provided opportunities to control mobile-phase 5

116

density and viscosity in such a way so that system pressure could be maintained only by controlling interface temperature. Strategies to modulate solvent density and viscosity to impart different pressure drops can be developed based on results reported in ref [17].

117

2.1.2

114 115

Pressure regulation through pumping

137

The configuration described in Figure 1(c) was proposed by Chester and Pinkston [18]. It has no active or passive mechanical means to control the pressure of the system. Full flow from the column first leads to the UV detector. Downstream to UV detector, a liquid organic solvent which is miscible to CO2 , is pumped and mixed with the SFC mobile-phase via a tee or mixer. Outlet to the mixer is directed to MS. Note that the MS-bound fluid should be passed through a flow restrictor to make the system work. The main benefit of this scheme is that the system volume is low, because no mechanical ABPR is employed, while the system pressure can still be independently controlled. Note that this additional stream may contain completely different set of solvents, compared to the SFC mobile-phase, which provides opportunity to introduce different solvents and additives to enhance MS signal, without affecting SFC separation process. On the downside, this system requires an additional pump providing pulse-free flow under constant-pressure mode. Also, this system does not allow to manipulate system pressure completely independent of any side-effects. Depending on the target pressure required volume of solvent addition will vary - which in turn changes analyte concentration to the MS. Concentration sensitive ionization techniques e.g. ESI (electro-spray ionization), may have problems in quantitation for this issue. Another problem is the process complexity, which may not be favorable in practical situations [2]. Application of this configuration was demonstrate by Pinkston et al [19] while screening a large library of pharmacuetically relevant compounds .

138

2.1.3

118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136

139 140 141 142 143

Full-flow through ABPR

Passing the full chromatographic flow through the ABPR has recently been possible with the introduction of low-volume ABPR design by major instrument providers like Shimadzu Corporation (Nexara UC) and Agilent Technologies Inc (Agilent 1260 SFC). Standard ABPR designs often add significant dead volume which can result into broad, tailing peaks leading to loss of chromatographic resolution. Introduction of low-volume ABPRs addressed this

6

164

issue. An additional technological challenge, however, can come from designing the transfer line from ABPR outlet to MS inlet [2]. Fully depressurized CO2 -rich solvent may not be able to transport all analytes to the MS, especially if the transfer line is too long. Also, low-volume of the ABPR may come as a design trade-off with the robustness. To address post-ABPR precipitation, a make-up flow is added which also improve MS signal. The make-up flow can be added either before or after the ABPR [20, 21]. Duval et al [21] compared the results between pre- and post-ABPR make-up flow addition and obtained similar quantitative results for atmospheric pressure chemical ionization (APCI). Several applications with full-flow ABPR has been reported in recent times. Hofstetter et al [22] used a Nexera UC system coupled to a Shimadzu LCMS-2020 single quad MS. The authors reported supercritical fluid extraction (SFE) of ketamine metabolites from dried urine for on-line quantification. Fassauer et al [23] had used the same system for chiral analysis of ketamine metabolites with antidepressant effects. Jiang et al [24] used a Nexera UC with an LCMS-8050 triple quad MS to determine alkylphenol ethoxylates in leafy vegetables. Wolrab et al [25] carried out SFC analyses of amino acids and related compounds with an Agilent 1260 SFC System which was directly coupled to a 6490 triple quad MS from Agilent. Bieber et al [20] carried out polarity-extended organic molecule screening in water samples employing an Agilent 1260 SFC system coupled with Agilent 6230 time-of-flight (TOF) MS. Laboureur et al [26] performed profiling of modified nucleosides from ribonucleic acid digestion by coupling an Agilent 1260 SFC system hyphenated to an Agilent 6540 Q-ToF mass spectrometer.

165

2.2

144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163

166 167 168 169 170 171 172 173 174 175

Split-flow introduction

In split-flow configuration a minor portion of the effluent from SFC column is diverted to MS inlet, whereas the major portion is passed through the ABPR. The main advantage is lesser constraints on the ABPR design. The main disadvantage, however, is that only a minor fraction of the column output is directed to MS. Flow-spliting configurations may affect MS response. E.g. for mass-flow sensitive ionization methods, like APCI, flow-split leads to lower response compared to the possibility of full flow to MS. For concentration dependent techniques e.g. ESI, however, loss in mass-flow is not a concern. Another issue that can lead to problems with quantitation is that the split ratio directed towards MS varies with mobilephase pressure or with any other changes that can vary the connector resistance vis-a-vis the main line resistance. Grand-Guillaume Perrenoud et al [27] reported an investigation on the 7

177

effect of the operating parameters on the split ratio and how that may affect MS output. In the following subsections two reported configurations under split-flow is discussed.

178

2.2.1

176

MS-split before UV

185

In this configuration (Figure 1(e)) a zero-volume tee is employed at the column outlet. A minor fraction of the flow (1 to 20%) [2] is diverted towards MS inlet and the other fraction is led through an UV detector to the ABPR. To ensure the minor flow to MS the connector should offer much stronger flow resistance compared to that offered by the ABPR. A useful part of this setup is the addition of another detector (UV) in the system, although neither detectors experience the full-flow coming out of the column. Zhao et al [28] reported this configuration that added a make-up pump to the MS-bound flow stream.

186

2.2.2

179 180 181 182 183 184

187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205

MS-split after UV with make-up pump

The configuration described in Figure 1(f) is followed by major SFC-MS instrument providers like Waters Corporation (Acquity UPC2 ) and Agilent Technologies Inc. In this configuration, the column outlet is first taken through a UV detector to a mixer. Flow of liquid organic make-up solvent(s), similar to configurations (c) and (d), is added to the mobilephase through this mixer. The flow from the mixer is then taken to a splitter where a minor fraction of the flow is led to the MS and the major fraction to the ABPR. This is a versatile configuration which provides the flexibility of managing mobile-phase decompression independent to the ABPR. But because the MS split is enacted through a passive splitter, there is no direct control on the solvent and analyte mass that is reaching the MS. For mass sensitive ionization techniques e.g. APCI, this can lead to problems in quantitation [27]. A solution to this problem could be to engage an active splitter but that may increase system volume. Addition of make-up flow helps later, during solvent decompression, but it also dilutes analyte concentrations. Depending on operating conditions this dilution factor may vary. Grand-Guillaume Perrenoud et al [27] estimated that under most common operating conditions of SFC-MS, fluctuation in dilution factor is between 1.1 to 1.5 when make-up flowrate is varied vis-a-vis the mobile-phase flowrate. Although there are early references [29], application of the configuration described in Figure 1(f) has significantly increased in recent times because of commercial availability. For example, Crepier et al [30] employed an Acquity UPC2 to hyphenate with an Ion Trap-

8

214

TOF MS from Shimadzu for fast pyrolysis characterization of bio-oils. Shi et al [31] used Acquity UPC2 hyphenated with Xevo G2-S Q-TOF and Vion IMS-QTOF from Waters Corp for mass profiling and comparison of lipidomes of various traditional Chinese medicines. Li et al [32] reported monitoring dissipation of thiacloprid in greenhouse vegetables and soil with Acquity UPC2 coupled with Xevo-TQD (Waters Corp). Borovcova et al [33] reported analysis of psychoactive substances in human urine employing Acquity UPC2 and Xevo TQS (Waters Corp). Prothmann et al [34] analyzed lignin-derived monomeric compounds in processed lignin samples with an Acquity UPC2 and Waters XEVO-G2 QTOF-MS (Waters Corp).

215

3

206 207 208 209 210 211 212 213

216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236

Managing solvent decompression

Managing solvent decompression inside SFC-MS connector tube is the most challenging and poorly understood part of SFC-MS coupling. Among the different contributors of peak broadening during the transfer process from column outlet to MS inlet, mobile-phase decompression inside the connector is the strongest contributor [27]. SFC mobile-phase is constituted of a high fraction of CO2 . Although CO2 does not play any direct role in ion formation [10], CO2 -rich mobile-phase is beneficial for MS because unlike liquid solvents CO2 does not need to be evaporated - rather it helps in better nebulization. CO2 -based mobile-phase expands to many times of its volume under atmospheric pressure. Smith et al [5] estimated that the masses of average clusters, formed when an SFC mobile-phase is expanding, are about 106 to 109 times less compared to the droplets formed during direct liquid introduction from an HPLC-MS interface. Although this comparison was done between neat CO2 and liquids, efficiency of CO2 with liquid co-solvent mixtures will be somewhere in between. Better nebulization process leads to better and faster desolvation of liquid solvent(s) that is mixed with CO2 . The expansion or decompression process, however, can lead to extremely poor MS signal if not managed properly. The most common problem is phase separation of mobile-phase inside the connector. Phase separation significantly increases possibility of analyte precipitation. Smith et al [6] observed that precipitation of analytes lead to ”spiking” signals. Although phase separation can greatly induce analyte precipitation, whether an analyte will finally precipitate or not, depends on its solubility in the liquid or vapor phases. E.g. even without phase separation certain analytes may precipitate on connector wall if they lose solubility

9

249

inside the depressurizing solvent. The fundamental factors that influence phase separation and analyte precipitation can be identified as (a) design and material of the connector - e.g. length and internal diameter(s) at various points along the length, and heat-transfer properties of connector material. And (b) pressure, temperature and composition of the solvent mixture inside the connector - before, during and after the decompression process. Design and material of the connector control the rate of decompression of the fluid and also determine the temperature along connector length. Pressures, temperatures and compositions of the fluid inside the connector, on the other hand, determine the phases and the physical properties e.g. solubility, viscosity etc. Although the motivation of this section is to present a detailed discussion on the roles of all these factors, a discussion on the causes and effects of phase separation inside connector is important to understand - which is presented first.

250

3.1

251

3.1.1

237 238 239 240 241 242 243 244 245 246 247 248

252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267

Phase separation inside the connector Factors influencing phase separation

A homogeneous phase separates in two or more phases if the pressure and/or the temperature is altered and the new condition does not support the fluid to stay in a single phase. For example when water temperature is taken to 100 o C at atmospheric pressure it turns into a liquid and a vapor phase through boiling. At lower pressures water boils at even lower temperatures. The combinations of pressures and temperatures, under which a pure compound starts boiling from liquid phase or starts condensing from vapor phase, constitutes a vapor liquid equilibrium (VLE) curve. For mixtures, phase separation can occur over a significantly higher pressure-temperature combinations. When phase separation occurs in a mixture and it separates in a liquid and vapor phase, both phases contain mixture components - but with different compositions. Only exception is an azeotropic mixture where both liquid and vapor compositions are the same [35]. Figure 2 demonstrates pressure-temperature conditions under which (a) neat CO2 , (b) various compositions of CO2 +methanol mixtures and (c) neat methanol, experience phase separation. The Vapor-Liquid Equilibrium curve of neat CO2 represents the infinite number of pressure-temperature combinations that lead to a liquid CO2 to boil or a vapor CO2 to condensate. The VLE curve of neat methanol presents similar pressure-temperature 10

285

information of methanol. The vapor-liquid two-phase region boundaries of CO2 +methanol mixtures encompass all the possible pressure-temperature combinations that lead to phase separation. Note from Figure 2 - with increasing proportion of methanol in the mixture, the pressure-temperature envelop of two-phase area first increases, reaches a maximum, and then decreases. More discussion on the utility of these phase boundaries are presented in section 3.4.2. Data used to plot these curves were obtained from REFPROP [36]. Further description of such phase diagrams can be obtained in reference [37]. Inside an SFC-MS connector, depending on the starting temperature and pressure, when a fluid is depressurized it may separate in two phases depending on the local temperature and pressure of the connector. For example (see in Figure 3), if mobile-phase pressure and temperature before depressurization is 100 bar and 50 o C, respectively, and the connector temperature, with the fluid in it, is also strictly maintained at 50 o C during depressurization - phase separation can be avoided for neat CO2 . Note from the figure that when pressure is reduced from 100 bar to atmospheric pressure at 50 o C, the conditions never pass through the VLE of neat CO2 (Figure 3). On the other hand, if the mobile-phase has 5% methanol in it, 50 o C cannot avoid phase separation, it has to be higher - e.g. 75 o C (Figure 3). With increasing composition of methanol in CO2 , increasing temperature is needed to avoid phase separation. More discussion on this topic is presented later in this section.

286

3.1.2

268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284

287 288 289 290 291 292 293 294 295 296 297 298 299

Flow behavior under phase separation

Behavior of a fluid flow while going through a phase separation can be very complex. Figure 4 presents example situations of conditions where, depending on the relative volume of vapor and liquid, the following types of flow are possible (a) Bubbly-flow: This flow pattern can be identified by the occurrence of mono-dispersed bubbles whose diameters are smaller than the channel diameter. Figure 4 (from ref [38]) shows occurrence of bubbly flow at the boundaries of a two-phase region - on the bubblepoint curve. In other words, such flow forms when pressure-temperature condition inside the connector just initiates boiling of the mobile-phase. The near-spherical shapes of the bubbles signify the dominance of surface tension of the vapor phase at the interface. The liquid volume inside the stream is much higher than the vapor volume. (b) Slug-flow: Under slug-flow conditions vapor phase moves as trains of elongated Taylor bubbles and are separated by liquid bodies. Figure 4 shows slug-flow conditions occurring deep inside the two-phase region. The vapor part occupies almost the entire channel cross

11

322

section and only a thin liquid film separates the bubbles from the wall. This flow happens when the vapor and the liquid volumes inside the connector are comparable. (c) Annular-flow: Annular flow occurs when the flow contains a high volume of vapor compared to the liquid. In Figure 4 annular flow is shown to be occurring at and near the dew-point curve. Under annular flow condition the liquid flows along the wall as an annular cylinder whereas the vapor flows as a cylinder, occupying the inner core. Figure 4 describes how the flow pattern can change inside the connector depending on the pressure-temperature conditions for a mixture. The figure presents a schematic plot of a two-phase region (can also be multiphasic, as shown in this figure), which has a critical point, a cricondenbar and a cricondentherm. Cricondenbar represents the maximum pressure point of the two-phase region whereas cricondentherm represents the maximum temperature. At points A, when the pressure-temperature conditions are outside the two-phase region, flow inside a connector is single homogeneous phase. As the temperature increases along a constant pressure from point A to B, (see Figure 4), or the pressure decreases at a constant temperature, the fluid just enters the two-phase region. Under such conditions it will experience a bubbly-flow because the liquid volume here is significantly more than the vapor volume. If the temperature is increased more at constant pressure, or the pressure is decreased at constant temperature, more vapor will form and the flow will be a slug-flow. At even higher temperatures or lower pressures, when along the respective constant-pressure and constant-temperature lines, the pressure-temperature combination moves more towards the dew curve, vapor formation will be very high and there will be an annular flow. If the initial condition is at the fully vapor side (point A at the right-hand side of dew curve), decreasing temperature will create the same flow pattern, but in opposite order.

323

3.1.3

300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321

324 325 326 327 328 329 330 331

Effects of phase separation

Visual representations of the effects of temperature and pressure on the fluid flow inside the connector, presented in Figure 4, can be helpful in understanding the effect of phase separation on the flow of analyte molecules. For all types of flows described above, both liquid and vapor part contains all the solvent compounds, but with different compositions. For example, for a CO2 +methanol mixture, both liquid and vapor part of the flow contain CO2 and methanol. Vapor is CO2 -rich whereas the liquid is methanol-rich. The exact composition depends on the original composition - when the mixture was in a single phase, and also on the local pressure and temperature.

12

340

Inside a two-phase flow analytes get distributed between both phases - similar to the physical process of a liquid-liquid extraction. As the physical properties are different for liquid and vapor phases, analytes traveling with liquid phase experience very different physical conditions compared to analytes traveling in the vapor phase. In addition, liquid and vapor travel at different velocities inside the connector depending on the slip ratio or velocity ratio between the two phases [39]. Velocity ratio depends on physical properties e.g. viscosity, surface tension, etc., and also on the net mass flow rate and connector geometry. This difference in flowrate between the two phases can be a potential source of band broadening, especially when an analyte is soluble in both phases.

341

3.2

332 333 334 335 336 337 338 339

342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363

Role of interface geometry

Connector geometry plays a pivotal role in the decompression process. A long connector, with a high length to diameter ratio (high aspect ratio), which leads to a continuous, gradual, linearly reducing pressure (and hence solvent density) inside the connector, is possibly the worst geometry. If phase separation occurs when depressurizing through such a connector, the deleterious effect will be imparted over a longer distance of the connector. As noted by Pinkston [2], for long connectors of length one meter or more, phase transition leads to segmented flow of CO2 -rich and modifier-rich phases, resulting in ”unstable, erratic, pulsating ESI spray and jagged peaks”. On the other hand, relatively shorter connectors with wider diameter but tapered or pinched at the end, likely serves the situation better [6]. In the following paragraphs a discussion on this topic is presented. Smith et al [6] summarized a list of connector geometries which were being applied as restrictors for a direct fluid injection setup for SFC-MS systems. Based on the design objectives, these geometries are still useful as connectors for modern SFC-MS interfacing both for split-flow and full-flow. Figure 5 lists schematic illustrations of a range of restrictor geometries. Figure 5(a) presents the cross-sectional view of a linear capillary restrictor. There are two sections of this design - the first section having a larger diameter compared to the second section. As long as the second section is not too long and the first section does not offer strong flow resistance, this design is a better option than having a long connector with short diameter. Cousin et al [8] described a short restrictor (3 mm x 5µm i.d.) which imparted a two-stage expansion with each stage heated to compensate for the expansion cooling and maintain analyte solubility. Figure 5(b) presents a drawn and tapered fused silica restrictor. The practical advantage of the drawn restrictor, compared to a similar length 13

364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397

restrictor but of small diameter capillary, is that the ”undrawn” portion of the capillary can facilitate effective heating to the fluid prior to expansion [6]. A major disadvantage of this type of restrictor is its fragility. The other restrictor designs are of shorter length with more favorable aspect ratio. The restrictor shown in Figure 5(c) is a polished ”integral” restrictor which is fabricated by polishing the closed end of a fused silica capillary till a small orifice of required diameter is obtained. Another approach involves internal deposition of a material to nearly close the end of a capillary, as shown in Figure 5(d). A fast drawing technique of fused silica can produce restrictors of similar dimensions (see Figure 5(e)). The design shown in Figure 5(f) is a diaphragm, or laser-drilled ”pinhole” orifice. Although it has some practical limitations, if properly manufactured can provide very efficient expansion process. The porous frit restrictor shown in Figure 5(g) resembles a column packed with sub-micron particles. Its geometry results in multiple fluidic paths with very high aspect ratios. The frit restrictor is robust in design and should not get easily plugged by particles-entrained in the fluid. The tortuous path offered by the restrictor, coupled with relatively long residence time facilitates efficient heat-transfer to the expanding fluid. This geometry improves transport of relatively volatile compounds, but may not be efficient for nonvolatiles. Actually very nonvolatile analytes may precipitate ultimately resulting in plugging. The last design (Figure 5(h)), the ”pinched” restrictor, is fabricated by crimping ca. 1 mm length of the termination of a segment of capillary platinum-iridium tubing (50-100 pm id.) The main advantage with these restrictors is the compatibility with higher temperatures and more polar fluids compared to fused silica restrictors. According to Smith et al [6], the main criterion while selecting a restrictor geometry should be the volatility of the analytes. Lesser volatile compounds benefit from low aspect ratio connector and operating conditions that maximize fluid density at the restrictor exit. For volatile compounds, the requirement is higher temperatures, and the designs which are the most rugged and can be conveniently replaced, are the more favorable ones. To arrest precipitation and to maintain a ”nucleation mode”, Smith [6] suggested to ensure that solvating conditions should be maintained till the end of the connector. This can be implemented by using short restrictors and maintaining fluid conditions that enhance solubility [6]. Similar view was published even earlier by Chester[40]. To stop condensation in a flame-ionization detector restrictor, and also to extend the molecular weight range of analysis, the author had suggested tapering the restrictor outlet, which helps keeping the pressure higher throughout the flow channel and imparts a more abrupt pressure gradient at

14

403

the outlet. More recently Petruzziello et al [41] reached very similar conclusion. The authors used a patented connector from Waters Corporation that delayed the decompression process in the connector line (see Figure 6). This modification from the original design significantly improved repeatability of analysis and restricted occurrence of spiky, choppy peaks (see Figure 7). A more detailed explanation on the usefulness of such connector design will be described in section 3.4 in conjunction with effects of temperature modulation.

404

3.3

398 399 400 401 402

405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429

Role of solubility and viscosity

Understanding the role of fluid composition and local pressure-temperature on physical properties e.g. analyte solubility, solvent viscosity etc. is important for having an insight regarding the events going on inside the connector. Figure 8 shows how solubility of a compound varies as a function of pressure and temperature. Although the solvent here is neat CO2 , similar behavior is also observed for solubility in solvent mixtures. Note in Figure 8, there is a pressure point ( 125 bar) above which increasing temperature at constant pressure increases solubility. This is the pressure inflection point. Below this point increasing temperature decreases solubility [42]. Solubilities of different compounds have different pressure inflection points, the value of which is generally higher than 100 bar for neat CO2 as a solvent. When there is phase separation, solubility in mixed-phase depends upon the respective compositions of liquid and vapor phases and analyte solubility in these two phases. Generally, relatively polar analytes should have more solubility in alcohol-rich liquid phase whereas nonpolar analytes should have more solubility in CO2 -rich vapor phase. If a relatively nonpolar liquid solvent is added in the make-up flow - nonpolar analytes will be more soluble in the liquid phase compared to depressurizing vapor phase which is losing its solvation power. Mobile-phase viscosity plays a critical role in determining the MS-bound solvent flowrate under a split-flow configuration (see section 2.2). Grand-Guillaume Perrenoud et al. [27] reported a comprehensive study demonstrating the effect of viscosity in determining the splitratio for a passive splitter and the resultant net flow of methanol inside the MS-bound solvent. Figure 9 demonstrates how mobile-phase flow rate and make-up pump flowrate influences the split-ratio and the dilution factor of the analytes. The mobile-phase composition was 90/10 (v/v,%) CO2 /methanol and make-up flow was neat methanol. Additional volume of methanol from the make-up flow dilutes analyte concentration in the MS-bound fluid - which is quantified as dilution factor in Figure 9. More discussion from this study is presented in 15

430

section 3.5.

431

3.4

432

3.4.1

433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460

Effects of controlling interface temperature Effects of heating the connector

Heating SFC-MS connector is being practiced for a long time. As observed by Chester [40], when the fluid temperature is increased, it improves MS signals of the peaks, starting with the early eluting ones. With increasing temperature, signals of even later eluting compounds start getting better. Smith [6] observed that if analytes precipitate inside the connector, they tend to get collected on connector wall. If the temperature at that local point inside the connector is made higher than analyte melting point, it can flow towards the connector end, and later transported to the MS, entrained inside high shear gas flow. A common misunderstanding regarding the utility of heating is that it increases analyte solubility in a depressurizing fluid which is fast losing its density and hence dissolving power. As explained in the previous section (section 3.3), increasing temperature does increase solubility, but only when the pressure is above the pressure inflection point. For a depressurizing fluid, usually under 100 bar, increasing temperature decreases solubility especially for analytes with low volatility. The real benefit that heating does is to help avoiding phase separation inside the connector. Smith et al [5] had suggested that for neat CO2 , to obtain good flow characteristics and also to avoid a two-phase separation during CO2 expansion, the fluid temperature before entering the connector should be at least 80 to 100o C. Figure 10 demonstrates the reason behind this suggestion. The figure shows isenthalpic (constant-enthalpy) curves of pure CO2 (red curves where the numbers written on it are the enthalpy value in kJ/g). The figure also shows constant density curves in blue (numbers on blue curves are density values in g/mL). Vapor Liquid Equilibrium curve of CO2 is shown as a solid green curve. On Figure 10, a black solid curve is drawn on the isenthalpic curve of 0.44 kJ/g (consult ref [43] for explanations on the information conveyed by these plots). This curve provides a boundary of what should be the minimum temperature of the fluid before the depressurization process starts, if one wants to avoid phase-separation during the decompression process. For example, if the pressure prior to the connector for a neat CO2 solvent is 150 bar, following the boundary of 0.44 kJ/g, the minimum starting temperature of MS-bound fluid should be 88 o C (see Figure 10). If the temperature is lesser than that, e.g. if the starting temperature is 75 o C, CO2 will 16

471

start condensing into liquid phase when the pressure drops down to 70 bar which makes the temperature drop down to 28 o (see the trajectory of the blue solid curve). Note that these predictions are based on the assumption of an isenthalpic decrease in temperature (consult ref [43] for details). In reality, if the connector temperatures (88 and 75 o C, respectively) are maintained throughout the length in isothermal condition, both temperatures can avoid phase separation (see the solid green straight lines in Figure 10). This also clarifies why it was suggested by Smith [6] that heating the fluid prior to expansion as well as heating the connector itself, improve and extend detection for compounds with low volatilities. Although the example provided here is with neat CO2 , similar observations hold for mixture mobile-phases as well - but with an additional criterion of expanding the two-phase region. Discussion in this direction is presented in the next section.

472

3.4.2

461 462 463 464 465 466 467 468 469 470

473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492

Is heating always better?

Although heating the connector can be useful because of the reasons discussed above, it may not always offer the best solution if the phase separation landscape is not properly understood. Note from Figure 2, and also from the discussion in section 3.1.1 - the pressuretemperature envelop that defines the two-phase regions of CO2 +methanol mixtures significantly increases in area with increasing amount of methanol in the mixture. Although after 50/50 (mol/mol,%) of CO2 /methanol mixture the area starts decreasing again, the cricondentherm keeps on increasing towards higher temperature with increasing methanol in the solvent. Similar trend can be observed with CO2 + other liquid co-solvent mixture. Another important observation that can be made from Figure 2 is - it takes increasingly higher temperature to avoid phase separation if the co-solvent concentration in the MS-bound solvent increases. This is the reason Chester[40] suggested to heat the connector as much as possible - e.g 400 o C provided the analytes are thermally stable. Figure 11 demonstrates possible consequences of choosing different temperatures when the mobile-phase is a CO2 +methanol mixture of 70/30 (mol/mol,%) composition. Assume that the starting point is 150 bar and 25 o C. This can be considered as a typical condition. The ABPR is often set at 150 bar and the pressure at the MS-split should not be very different from this pressure (see Figure 1(f)). Also, by the time the mobile-phase reaches the MS-split the temperature almost equilibrates to the room temperature. CO2 +methanol mixture of 70/30 (mol/mol,%) can be considered as a typical composition, especially with an addition of make-up fluid.

17

493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526

Note from Figure 11, if the SFC-MS interface is kept at room temperature, the MS-bound fluid decompresses till around 50 bar as a single phase. Decompression till this point does not lead to any significant change in temperature because of the near vertical isenthalpic curve of 0.16 kJ/g (consult ref [43] for detailed explanation). Further decompression, however, takes the fluid inside the two-phase region leading to phase separation. If a linear decrease in the pressure is assumed, at room temperature, approximately the last one-third of the connector length will experience the consequences of phase separation. If, on the other hand, the temperature is increased to 75 o C before decompression, the solvent will reach two-phase region at a higher pressure - at around 100 bar. In this case, the solvent flow will experience phase separated flow over two-third fraction of the connector length, which certainly should worsen the situation. With the solvent of CO2 +methanol mixture of 70/30 (mol/mol,%) composition, to effectively avoid phase separation, the interface needs to be heated to atleast 150 o C and maintain that temperature throughout the connector length, provided the analytes do not degrade at that temperature. If isenthalpic cooling of the solvent is allowed during decompression process, the solvent should be heated atleast upto 200 o C. Instead of heating, another approach to avoid the two-phase region can be cooling the fluid to make it as much liquid-like as possible [6]. Note from Figure 11 that if the temperature is cooled down to 0 o C, the fluid can depressurize till 25 bar from 150 bar before entering the two-phase region. In this case, inside a connector tube, phase-separation will occur only during the last one-sixth of the connector length. If the temperature is brought down even further, e.g. -30 o C, the fluid can depressurize till 12 bar before entering the two-phase region. Which means less than one-twelfth part of the connector length will experience phase separation. If low temperature operations are carried out with a connector that can create a sharp pressure drop at the end with a tapered-tip or pinched-tip (see section 3.2), any phase separation inside the connector can be completely avoided. Figure 11 also demonstrates the utility and best operating temperatures of connectors which are designed for high pressure-drop over a small length e.g. tapered-tip connectors (section 3.2). If the fluid temperature before depressurization is high, e.g. 75 o C, the taperedtip should be designed for a very high pressure drop, e.g 90 bar. Such high pressure drop requirement increases design challenges and also leads to faster wearing of tip material. At lower temperatures, on the other hand the tip needs to enforce a lower pressure drop. In addition, at lower temperatures viscosity is higher - which enforces high pressure drop even with lower volume of make-up fluid. Note from Figure 2, there is no significant difference

18

538

between the boundaries of 5% methanol and 50% methanol in CO2 +methanol mixture at 25 o C. This explains the observation made by Petruzziello et al [41] that the improved performance gained employing a tapered-tip type device could be maintained even at reduced make-up flow rates. Another advantage of cooling the connector is - if the solvent is cooled before the depressurization starts, even with significant depressurization, there is only minor change in solvent density. Note the changes in constant-density curves (blue curves) in Figure 11. At lower temperatures solvent density reduction is negligible even with high pressure drop. At higher temperatures, on the other hand, similar range of depressurization leads to a 4 to 7 fold drop in fluid density. Analytes which are soluble only at high density mobile-phase may start precipitating even when a phase separation is avoided. This situation will not arise if the mobile-phase is cooled.

539

3.5

527 528 529 530 531 532 533 534 535 536 537

540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558

Effects of interface solvent composition

The final solvent constituents and composition inside the connector is determined by the flowrates, constituents and compositions of the SFC mobile-phase and those of the make-up solvent(s). Make-up flow is an ubiquitous feature of current SFC-MS operations. Addition of liquid organic solvent(s) after the chromatographic column is done for various purposes mainly to avoid analyte precipitation on the connector wall. Although solvent additives are added to enhance MS signal, the discussions presented in this section are mostly confined to solvent composition role in avoiding analyte precipitation. Addition of make-up solvent can significantly alter solvent composition inside the connector. E.g. if the SFC mobile-phase flowrate is 1 mL/min and the composition is 85/15 (vol/vol,%) of CO2 /methanol, addition of 0.3 mL/min of methanol as make-up solvent alters the composition to around 65/35 (vol/vol,%) of CO2 /methanol (mass or mole compositions should be considered for more accurate estimation). If in the same mobile-phase, 0.5 mL/min of make-up flow is added in place of 0.3 mL/min, the altered composition will be approximately 57/43 (vol/vol,%) of CO2 /methanol. Referred to the discussions presented in sections 3.1.2 and 3.4.1, we may note that increasing liquid solvent composition in the MS-bound fluid may not always help in avoiding phase separation - however, it influences the flow behavior inside the connector. E.g. increasing make-up flow may alter an annular flow to a slug-flow or bubbly-flow. Depending on the composition and constituent of the fluid inside the connector, certain analytes may be agnostic to the change in flow behavior, for some 19

559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592

other analytes it can be crucial. Relationship between the make-up flow rate and the intensity and repeatability of MS response is not straight forward and no particular trend is detected. Akbal and Hopfgartner [44] found no effect of make-up flowrate on the MS response for compounds like theobromine, metoprolol, bosentan etc. when the flow was increased from 0 to 0.7 mL/min. Duval et al [21] reported increasing response with increasing make-up flowrate for triacylglycerols and hydroxylated triacylglycerols (see Figure 12). For fatty acids, e.g. linoleic acid, the response decreased with increasing make-up flow. There are reports [10, 45] that demonstrated existence of an optimum flowrate or optimum percent of liquid organic co-solvent, in the MS-bound fluid flow, that performs the best. Sadoun et al [10] noted that for a full flow SFC-MS system (with no make-up flow) the ion formation optimum was obtained at 2-3% of polar modifier in the mobile-phase in one example, employing a flowrate of 1 mL/min, and 20% of modifier in another example with flowrate of 0.15 mL/min. The authors noted that despite the apparent difference in optimum conditions, 20% of a 0.15 mL/min flow and 2-3% of a 1 ml/min flow, both roughly result into 0.020-0.030 mL/min of polar modifier to the electrospray region. This report did not contain any repeatability study. Fujito et al [45] reported the effect of varying make-up flow rate in a full flow SFC-MS system (see Figure 1(d)). The authors introduced four different make-up flowrates - 0.05, 0.1, 0.2, and 0.4 mL/min of methanol, in a mobile-phase flow of 0.6 mL/min where the co-solvent percent increased from 2 to 80%. Among the three flow rates, authors obtained the highest signal intensity at 0.1 mL/min of make-up flow (Figure 13). Repeatability of analysis was similar for flowrates of 0.1, 0.2, and 0.4 mL/min. Only at 0.05 mL/min repeatability was poor. Reason for the effectiveness of a lower make-up flow (0.1 mL/min) was assumed to be resulting from an increase in the desolvation efficiency, although even lower flowrates (e.g. 0.05 mL/min) were not helpful for repeatability problems. The authors further calculated the total fluid flowrate and composition of the MS-bound solvent and investigated the relationship (if any) between repeatability and co-solvent flow rate. They detected that when the make-up solvent flow rate was 0.05 mL/min, the total co-solvent flow rate in MS-bound fluid was below 0.1 mL/min, during the entire time when almost all the compounds eluted. The authors also found that repeatability could be improved with increasing co-solvent flow rate as long as total co-solvent flow was not more than 0.4 mL/min. Based on their observations they noted that total co-solvent flow should be more than 0.1 mL/min to keep the ESI spray stable. An optimum flow lies between 0.10.2 mL/min for improving both sensitivity and repeatability. Grand-Guillaume Perrenoud et al [27] conducted a detailed study on estimating the net

20

613

co-solvent flowrate in the MS-bound flow when a split is employed to interface with the MS (see Figure 1(d)). They estimated that the net co-solvent (methanol) flowrate to MS was between 0.125 to 0.3 mL/min under typical operating conditions which are applied for both isocratic and gradient modes in SFC-MS. They defined typical operating conditions as - SFC flow rate between 1.0 to 3.0 mL/min, flowrate of make-up fluid between 0.3 to 0.6 mL/min, ABPR pressure set between 120 to 150 bar and mobile-phase composition varying between 2 to 40% of MeOH in CO2(v/v). The authors noted that such flow range of methanol in MS-bound fluid is sufficient for preventing analyte precipitation during solvent decompression and also to ensure both optimal ESI spray formation and good proton transfer during ionization process. Phase diagrams can be consulted to develop a better insight into the process of adding extra co-solvent to the mobile-phase. Chester [46] had explained the role of adding make-up solvent to the MS-bound stream in a setup where a pressure regulator is replaced with a tee that adds make-up flow to the main stream (see Figure 1(c)). In this case mobile-phase composition is changed isothermally (from point 2 to 4 in Figure 14). The main purpose of adding a make-up flow here was to increase the viscosity of the fluid. This helped in keeping the pressure high inside the connecting tube. This arrangement could prevent phase separation inside the connector, or at least delay it till the flow reaches close to the tube end. Similar strategy can be extended along with the cooling strategy explained in section 3.4.2. For a tapered-tip connector, if the pressure drop is not high enough across the tip, solvents of higher viscosity could be connected to avoid phase separation inside the connector.

614

4

593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612

615 616 617 618 619 620 621 622 623

Conclusion

This review highlights the main challenges to design and develop proper methods for interfacing an SFC system with MS. The most challenging problem in SFC-MS interfacing is to manage solvent depressurization while maintaining chromatographic resolution. Solvent depressurization often leads to phase separation inside the connector which may result in analyte precipitation and poor MS response if not properly handled. There are two different practices in SFC-MS that tries to counter the problems related to phase separation - (a) heating the interface connector and (b) add a liquid solvent stream (make-up fluid) to the SFC outlet to make the solvent more liquid-like before entering the MS connector. This review demonstrates that although heating helps in avoiding phase separation in-

21

624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639

side the connector, different solvent compositions has different temperature requirements. Without any thermodynamic data of that fluid mixture, it could be hard to estimate how high the connector should be heated. Another possible way to avoid phase separation could be keeping the connector at room temperature and design the connector for a more favorable decompression scheme. Connectors that allow late decompression, preferably delaying decompression till the end of the connector tube, helps avoiding or minimizing analyte precipitation on connector wall. Such connector designs may perform better if the solvent is cooled further. Role of make-up fluid in influencing MS response is more complex and depends on the nature of the analytes. Based on the liquid versus CO2 composition inside the MS-bound fluid - the phase-separated flow may occur as (a) bubbly-flow when the liquid part is significantly higher, (b) slug-flow when the liquid part and CO2 are comparable, and (c) annular-flow when the CO2 part is significantly higher. The make-up flow, contributing a new stream of liquid solvent, not only contributes in increasing the liquid part and hence the flow behavior, it also modify solvent properties - e.g. polarity, which can result in different responses for different groups of analytes.

22

640

Acknowledgment

642

Helpful discussion with Jason Hill, Dr Micheal Fogwill and Dr Marian Twohig of Waters Corporation gratefully acknowledged.

643

References

641

644 645

646 647 648

649 650

651 652 653

654 655

656 657

658 659

660 661 662

663 664 665 666

[1] M. T. Combs, M. Ashraf-Khorassani, L. T. Taylor, Packed column supercritical fluid chromatography-mass spectroscopy: A review, J. Chromatogra. A 785 (1997) 85–100. [2] J. D. Pinkston, Advantages and drawbacks of popular supercritical fl uid chromatography mass-spectrometry interfacing approachesa users perspective, Eur. J. Mass Spectrom. 11 (2005) 189–197. [3] L. G. Randall, A. L. Wahrhaftig, Dense gas chromatograph/mass spectrometer interface, Anal. Chem. 50 (1978) 1703–1705. [4] R. D. Smith, J. C. Fjeldsted, M. L. Lee, Direct fluid injection interface for capillary supercritical fluid chromatography-mass spectrometry, J. Chromatogra. 247 (1982) 231– 243. [5] R. D. Smith, H. R. Udseth, Mass spectrometry with direct supercritical fluid injection, Anal. Chem. 55 (1983) 2266–2272. [6] R. D. Smith, H. T. Kalinoski, H. R. Udseth, Fundamentals and practice of supercritical fluid chromatography-mass spectrometry, Mass Spec. Rev. 6 (1987) 445–496. [7] P. J. Arpino, J. Cousin, J. Higgins, Supercritical fluid chromatography-mass spectrometry coupling, Tr. Anal. Chem. 6 (1987) 69–73. [8] J. Cousin, P. J. Arpino, Construction of a supercritical fluid chromatograph-mass spectrometer instrument system using capillary columns, and a chemical ionization source accepting high flow-rates of mobile phase, J. Chromatogra. A 398 (1987) 125–141. [9] P. J. Arpino, F. Sadoun, H. Virelizier, Reviews on recent trends in chromatography/mass spectrometry coupling. Part IV. reasons why supercritical fluid chromatography is not so easily coupled with mass spectrometry as originally assessed, Chromatographia 36 (1993) 283–284. 23

667 668 669

670 671

672 673 674

675 676 677 678

679 680 681 682

683 684 685

686 687

688 689 690 691

692 693 694

[10] F. Sadoun, H. Virelizier, P. J. Arpino, Packed-column supercritical fluid chromatography coupled with electrospray ionization mass spectrometry, J. Chromatogra. 647 (1993) 351–359. [11] P. J. Arpino, P. Haas, Recent developments in supercritical fluid chromatography-mass spectrometry coupling, J. Chromatogra. A 703 (1995) 479–488. [12] H. T. Kalinoski, L. O. Hargiss, Design and industrial applications of a removable probe interface for direct capillary supercritical-fluid chromatography-mass spectrometry, J. Chromatogra. 474 (1989) 69–82. [13] P. Sjoberg, K. E. Markides, Capillary column supercritical fluid chromatographyatmospheric pressure ionisation mass spectrometry interface performance of atmospheric pressure chemical ionisation and electrospray ionisation, J. Chromatogr. A 855 (1999) 317–327. [14] D. G. Morgan, K. L. Harbol, J. N. P. Kitrinos, Optimization of a supercritical fluid chromatographatmospheric pressure chemical ionization mass spectrometer interface using an ion trap and two quadrupole mass spectrometers, J. Chromatogra. A 800 (1998) 39–49. [15] S. Hoke, J. Tomlinson, R. Bolden, K. Morand, J. Pinkston, K. Wehmeyer, Increasing bioanalytical throughput using pcsfc-ms/ms: 10 minutes per 96-well plate, Anal. Chem. 73 (2001) 3083–3088. [16] M. Hayward, Q. Han, Interfacing mass spectrometry (MS) to chromatography: Whats next?, Presentation - 45th Anniversary of American Laboratory (2013). [17] A. Tarafder, K. Kaczmarski, D. P. Poe, G. Guiochon, Use of the isopycnic plots in designing operations of supercritical fluid chromatography. V. Pressure and density drops using mixtures of carbon dioxide and methanol as the mobile phase, J. Chromatogr. A 1258 (2012) 136–151. [18] T. Chester, J. Pinkston, Pressure-regulating fluid interface and phase behavior considerations in the coupling of packed-column supercritical fluid chromatography with lowpressure detectors, J. Chromatogr. 807 (1998) 265.

24

695 696 697

698 699 700

701 702 703 704 705

706 707 708 709

710 711 712 713

714 715 716 717 718

719 720 721 722

723 724

[19] J. D. Pinkston, D. Wen, K. L. Morand, D. A. Tirey, D. T. Stanton, Comparison of LC/MS and SFC/MS for screening of a large and diverse library of pharmaceutically relevant compounds, Anal. Chem 78 (1998) 7467–7472. [20] S. Bieber, G. Greco, S. Grosse, T. Letzel, Rplc-hilic and sfc with mass spectrometry: Polarity-extended organic molecule screening in environmental (water) samples, Anal. Chem. 89 (2017) 7907–7914. [21] J. Duval, C. Colas, V. Pecher, M. Poujol, J.-F. Tranchant, E. Lesellier, Hyphenation of ultra high performance supercritical fluidchromatography with atmospheric pressure chemical ionisation highresolution mass spectrometry: Part 1. study of the couplingparameters for the analysis of natural non-polar compounds, J. Chromatogr. A 1509 (2017) 132–140. [22] R. Hofstetter, G. Fassauer, A. Link, Supercritical fluid extraction (SFE) of ketamine metabolites from dried urine and on-line quantification by supercritical fluid chromatography and single mass detection (on-line SFE SFC MS), J. Chromatogr. B 1076 (2018) 77–83. [23] G. Fassauer, R. Hofstetter, M. Hasan, S. Oswald, C. Mode, W. Siegmund, A. Link, Ketamine metabolites with antidepressant effects: Fast, economical, and eco-friendly enantioselective separation based on supercritical-fluid chromatography (sfc) and single quadrupole ms detection, J Pharm Biomed Anal 146 (2017) 410–419. [24] Z.-J. Jiang, X.-L. Cao, H. Li, C. Zhang, A. Abd El-Aty, F. Jin, H. Shao, M.-J. Jin, S.-S. Wang, Y.-X. She, J. Wang, Fast determination of alkylphenol ethoxylates in leafy vegetables using a modified quick, easy, cheap, effective, rugged, and safe method and ultra-high performance supercritical fluid chromatographytandem mass spectrometry, J Chromatogra A 1525 (2017) 161–172. [25] D. Wolrab, P. Frhauf, C. Gerner, Direct coupling of supercritical fluid chromatography with tandem mass spectrometry for the analysis of amino acids and related compounds: Comparing electrospray ionization and atmospheric pressure chemical ionization, Anal. Chem. Acta 981 (2017) 106–115. [26] L. Laboureur, N. Bonneau, P. Champy, A. Brunelle, D. Touboul, Structural characterisation of acetogenins from annona muricata by supercritical fluid chromatography

25

725 726

727 728 729 730

731 732 733

734 735 736

737 738 739 740

741 742 743 744 745

746 747 748 749

750 751 752

753 754

coupled to high-resolution tandem mass spectrometry, Phytochem Anal 28 (6) (2017) 512–520. [27] A. G.-G. Perrenoud, J.-L. Veuthey, D. Guillarme, Coupling state-of-the-art supercritical fluid chromatography andmass spectrometry: From hyphenation interface optimization tohigh-sensitivity analysis of pharmaceutical compounds, J. Chromatogra. A 1339 (2014) 174–184. [28] Y. Zhao, G. Woo, S. Thomas, D. Semin, P. Sandra, Rapid method development for chiral separation in drug discovery using sample pooling and supercritical fluid chromatography-mass spectrometry, J. Chromatogr. A 1003 (1-2) (2003) 157–166. [29] M. Garzotti, L. Rovatti, M. Hamdan, Coupling of a supercritical fluid chromatography system to a hybrid (Q-TOF 2) mass spectrometer: on-line accurate mass measurements, Rapid Comm. Mass Spec. 15 (2001) 1187–1190. [30] J. Crepier, A. L. Masle, N. Charon, F. Albrieux, P. Duchene, S. Heinisch, Ultra-high performance supercritical fluid chromatography hyphenated to atmospheric pressure chemical ionization high resolution mass spectrometry for the characterization of fast pyrolysis bio-oils, J. Chromatogr. B 1086 (2018) 38–46. [31] X. Shi, W. Yang, S. Qiu, J. Hou, W. Wu, D. Guo, Systematic profiling and comparison of the lipidomes from panax ginseng, p. quinquefolius, and p. notoginseng by ultrahigh performance supercritical fluid chromatography/high-resolution mass spectrometry and ion mobility-derived collision cross section measurement, J. Chromatogr. B 1548 (2018) 64–75. [32] R. Li, Z. Chen, F. Dong, J. Xu, X. Liu, X. Wu, X. Pan, Y. Tao, Y. Zheng, Supercritical fluid chromatographic-tandem mass spectrometry method for monitoring dissipation of thiacloprid in greenhouse vegetables and soil under different application modes, J. Chromatogr. B 1081-1082 (2018) 25–32. [33] L. Borovcova, V. Pauk, K. Lemr, Analysis of new psychoactive substances in human urine by ultra-high performance supercritical fluid and liquid chromatography: Validation and comparison, J Sep SciArticle in Press (2018). [34] J. Prothmann, M. Sun, P. Spgel, M. Sandahl, C. Turner, Ultra-high-performance supercritical fluid chromatography with quadrupole-time-of-flight mass spectrometry 26

755 756

757 758

759 760

761 762 763

764 765 766

767 768

769 770 771

772 773 774

775 776 777

778 779 780

781 782 783

(UHPSFC/QTOF-MS) for analysis of lignin-derived monomeric compounds in processed lignin samples, Anal. Bioanal. Chem. 409 (30) (2017) 7049–7061. [35] Vapor - liquid equilibrium (Accessed: April 2018). URL https://en.wikipedia.org/wiki/Vapor-liquid equilibrium [36] E. Lemmon, M. Huber, M. McLinden, D. Friend, National Institute of Standards and Technology Standard Reference Data Program (2012) Gaithersburg 9.0.

[37] Fluid phase behavior: Pressure-temperature diagram (P-T diagram) (Accessed: April 2018). URL perminc.com/resources/fundamentals-of-fluid-flow-in-porous-media/chapter-5-misci [38] B. Pinho, S. Girardon, F. Bazer-Bachi, G. Bergeot, S. Marre, C. Aymonier, A microfluidic approach for investigating multicomponent system thermodynamics at high pressures and temperatures, Lab Chip 14 (2014) 3843–3849. [39] Slip ratio (gasliquid flow) (Accessed: April 2018). URL https://en.wikipedia.org/wiki/Slip ratio (gas-liquid flow) [40] T. L. Chester, Capillary supercritical-fluid chromatography with flame-ionization detection: reduction of detection artifacts and extension of detectable molecular weight range, J. Chromatogra. 299 (1984) 424–431. [41] F. Petruzziello, A. G.-G. Perrenoud, A. Thorimbert, M. Fogwill, S. Rezzi, Quantitative profiling of endogenous fat-soluble vitamins and carotenoids in human plasma using an improved uhpsfc-esi-ms interface, Anal. Chem 89 (2017) 7615–7622. [42] K. D. Bartle, A. A. Clifford, S. A. Jafar, Measurement of solubility in supercritical fluids using chromatographic retention: the solubility of fluorene, phenanthrene, and pyrene in carbon dioxide, J. Chem. Eng. Data 35 (1990) 355–360. [43] A. Tarafder, P. Iraneta, G. Guiochon, K. Kaczmarski, D. P. Poe, Estimations of temperature deviations in chromatographic columns using isenthalpic plots. I. Theory, J. Chromatogr. A 1366 (2014) 126–135. [44] L. Akbal, G. Hopfgartner, Effects of liquid post-column addition in electrospray ionizationperformance in supercritical fluid chromatographymass spectrometry, J. Chromatogr. A 1517 (2017) 176–184. 27

784 785 786

787 788 789

790 791 792 793

[45] Y. Fujito, Y. Hayakawa, Y. Izumi, T. Bamba, Importance of optimizing chromatographic conditions and massspectrometric parameters for supercritical fluidchromatography/mass spectrometry, J. Chromatogra. A 1508 (2017) 138–147. [46] T. L. Chester, Determination of pressuretemperature coordinates of liquidvapor critical loci by supercritical fluid flow injection analysis, J. Chromatogra. A 1037 (2004) 393– 403. [47] A. Tarafder, G. Guiochon, Use of isopycnic plots in designing operations of supercritical fluid chromatography: II. The isopycnic plots and the selection of the operating pressuretemperature zone in supercritical fluid chromatography, J. Chromatogr. A 1218 (2011) 4576–4585.

28

794

795

Figure captions 1

796 797 798 799 800 801 802 803 804 805 806 807 808 809 810

2

811 812 813 814 815 816 817 818 819 820 821 822 823 824

3

Schematic representations of various SFC-MS interfacing configurations. (a) A chromatographic column is connected directly to an MS through a pressure restrictor where the end of the restrictor is subjected to heating, (b) same as in (a) but here the restrictor is subjected to heating at the beginning of the restrictor, (c) here flow from the column is first passed through a UV detector, which is then mixed with a fixed-pressure pump flow, before being expanded to MS through a restrictor, (d) here flow from the column is passed through an automated back-pressure regulator (ABPR) before the MS and the flow is mixed with a make-up pump flow, either before or after the ABPR, (e) here a part of the outlet from the column is directed towards an MS whereas the other part is taken to an ABPR passing through a UV detector, (f) in this configuration, flow from the column is first passed through a UV detector, which is then mixed with a make-up pump flow. A part of the mixed stream is then taken to the MS and the other part to an ABPR. Detailed description of these configurations are provided in section 2. Vapor Liquid Equilibrium curves of neat CO2 and neat methanol, and twophase regions of different CO2 +methanol (mol/mol,%) mixtures. The figure shows that with increasing amount of methanol in CO2 , the two-phase region first increases and then decreases. The vapor-liquid equilibrium data were obtained from REFPROP [36] to plot these curves. Note that REFPROP could not converge solutions for all the solutions - so parts of the two-phase regions of 70/30, 50/50 and 30/70 CO2 /methanol mixtures were hand-drawn to present the complete boundary. Effect of mobile-phase composition on phase separation inside SFC-MS connector. If the mobile-phase is 100 % CO2 and the entire connector temperature, with the decompressing fluid in it, is maintained at 50 o C, there will not be any phase separation. On the other hand, if the composition is 95/05 (mol/mol,%) CO2 /methanol, at 50 o C there will be phase separation for the decompressing fluid. The vapor-liquid equilibrium data were obtained from REFPROP [36].

29

32

33

34

825

4

826 827 828 829 830 831 832

5

833 834 835

6

836 837 838

7

839 840 841

8

842 843

9

844 845 846 847 848 849 850 851 852

10

Schematic representation of a two-phase region based on Figure 3a of ref [38]. Note that the critical point separates the so-called bubble-point curve with the dew-point curve. At the bubble-point curve a single phase liquid mixture starts boiling into a liquid and vapor phase. At the dew-point curve a single phase vapor mixture starts condensing to form liquid droplets. Inside the two-phase region, at different pressure-temperature combinations the relative amount of vapor and liquid will be different. More description in section 3.1.2. Cross section view (not to scale) of different restrictor designs (from ref [6]) employed for SFC-MS interfacing. Discussions on these interfaces are presented in section 3.2. Schematic representation of the design differences between (A) a commercially available interface, and (C) a Waters Corporation prototype interface. Figure (B) shows pictures of the actual interfaces. From ref [41]. (A) Spiky or jagged peaks resulting from inconstant spray pulsing because of possible phase separation. (B) Better peak shapes using the prototype connector discussed in Figure 6. From ref [41]. Solubility data of phenanthrene as a function of pressure at different temperatures in neat CO2 . Data obtained from ref [42]. (A) Estimated values of split ratio and (B) dilution factor of the mobile-phase caused by the mobile-phase flowrate and the sheath (make-up) flowrate. Composition of the mobile-phase was kept fixed at 90/10 (v/v,%) of CO2 /methanol. ABPR was maintained at 150 bar. From ref [27] Plots of isenthalpic or constant-enthalpy (in red) and isopycnic or constantdensity (in blue) curves of pure CO2 on pressure-temperature plane. The values in red have kJ/g as the dimension and that in blue have g/mL. Data to generate these curves were obtained from REFPROP [36]. For more descriptions of the usage of these curves consult references [43, 47].This figure is used to explain the effect of connector heating/cooling in section 3.4.1.

30

35

36

37

38 39

40

41

853

11

854 855 856 857 858 859 860

12

861 862 863 864 865

13

866 867 868 869 870 871 872 873

14

Plots of isenthalpic or constant-enthalpy (in red) and isopycnic or constantdensity (in blue) curves of 70/30 (mol/mol,%) CO2+methanol mixture on pressure-temperature plane. The values in red have kJ/g as the dimension and that in blue have g/mL. Data to generate these curves were obtained from REFPROP [36]. For more descriptions of the usage of such curves consult references [43, 47]. This figure is used to explain the effect of connector heating/cooling in section 3.4.1. Demonstration of the differences in variations of MS response as a function of make-up flowrates. For the hydroxylated triacylglycerols, MS response increased with increasing flowrates. For fatty-acids, on the other hand, the response decreased. For some analytes the response was not affected by flowrates. Consult ref. [21] for more details. Comparison between normalized signal intensity of 441 compounds, conducted with three different flow rates of make-up solvent - 0.05, 0.1, and 0.4 mL/min. Results show that the highest response was achieved with 0.1 mL/min flowrate. Consult ref [45] for more details. Effects of changing temperature and composition for a two-component mixture. Starting from point (1), with a temperature of 75 o C and pressure of 150 bar, the method changes the temperature to 25 o C and then takes the composition to 75 %mol of methanol in CO2 . Depressurization process at this condition is shown by point (5). Consult ref [46] for more details.

31

42

43

44

45

(a)

MS

Restrictor

(b)

UV

Restrictor

MS

Pump

Full-flow introduction

(c) MS

UV

(d)

ABPR

Pump

MS

Pump

MS

(e) UV

ABPR

Split-flow introduction

MS

(f) ABPR

UV

Pump

Figure 1: Schematic representations of various SFC-MS interfacing configurations. (a) A chromatographic column is connected directly to an MS through a pressure restrictor where the end of the restrictor is subjected to heating, (b) same as in (a) but here the restrictor is subjected to heating at the beginning of the restrictor, (c) here flow from the column is first passed through a UV detector, which is then mixed with a fixed-pressure pump flow, before being expanded to MS through a restrictor, (d) here flow from the column is passed through an automated back-pressure regulator (ABPR) before the MS and the flow is mixed with a make-up pump flow, either before or after the ABPR, (e) here a part of the outlet from the column is directed towards an MS whereas the other part is taken to an ABPR passing through a UV detector, (f) in this configuration, flow from the column is first passed through a UV detector, which is then mixed with a make-up pump flow. A part of the mixed stream is then taken to the MS and the other part to an ABPR. Detailed description of these 32 configurations are provided in section 2.

Figure 2: Vapor Liquid Equilibrium curves of neat CO2 and neat methanol, and two-phase regions of different CO2 +methanol (mol/mol,%) mixtures. The figure shows that with increasing amount of methanol in CO2 , the two-phase region first increases and then decreases. The vapor-liquid equilibrium data were obtained from REFPROP [36] to plot these curves. Note that REFPROP could not converge solutions for all the solutions - so parts of the two-phase regions of 70/30, 50/50 and 30/70 CO2 /methanol mixtures were hand-drawn to present the complete boundary.

33

Figure 3: Effect of mobile-phase composition on phase separation inside SFC-MS connector. If the mobile-phase is 100 % CO2 and the entire connector temperature, with the decompressing fluid in it, is maintained at 50 o C, there will not be any phase separation. On the other hand, if the composition is 95/05 (mol/mol,%) CO2 /methanol, at 50 o C there will be phase separation for the decompressing fluid. The vapor-liquid equilibrium data were obtained from REFPROP [36].

34

Cricondenbar

A Critical point

B

Cricondentherm

D A

Figure 4: Schematic representation of a two-phase region based on Figure 3a of ref [38]. Note that the critical point separates the so-called bubble-point curve with the dew-point curve. At the bubble-point curve a single phase liquid mixture starts boiling into a liquid and vapor phase. At the dew-point curve a single phase vapor mixture starts condensing to form liquid droplets. Inside the two-phase region, at different pressure-temperature combinations the relative amount of vapor and liquid will be different. More description in section 3.1.2.

35

of a heated region of fused silica tubing from which the polyimide coating (shown as dark lines in Fig. 10) has been removed. Drawn capillaries or other variations (E) resulting in larger effective aspect ratios, typically in the range of 102-105, can be easily fabricated and are less likely to plug than straight walled capillaries (27,31). This also includes a commercially available variation (Suprex) that is formed by compression with heating near the end A

r

1

I

I

I

I

Figure 10. Cross section diagram (not to scale) of various restrictor designs Figure 5: Cross section view (not to scale) of different restrictor designs (from ref [6]) emSFC. All restrictors the inpinhole laserwhich have been explored ployed for SFC-MS interfacing. for Discussions on these interfaces except are presented section 3.2. drilled orifice (F) and the pinched restrictor (H) are fabricated from fused silica capillary tubing coated witb noliwnide (dark lines). See Table I1 for detailed descriptions. 36

Spiked peak profiles can induce quantitation issues and will often result in large RSD values. Although altering the make-up fluid flow rate could help minimize spray pulsing, flow rate increase was not considered due the increased solvent consumption and to an observed decrease in sensitivity at high make-up flow rates. Another option consists of minimizing the post-decompression volume. In other words forcing the decompression to occur as close as possible to the ESI spray. Since the commercially available UHPSFC-MS interface (Figure 4A)

Figure 4. UHPSFC-MS interface design schematically detailing the

Figure 6: Schematic representation of the design differences between (A) a commercially fluidic differences theCorporation commercially available available interface, and between (C) a Waters prototype interface.interface Figure (B)(A) shows pictures of the actual interfaces. From (C). ref [41].Picture (B) showing the fluidic and the prototype interface

domain of corresponding interfaces removed from the ESI probe assembly.

contains considerable post-decompression volume (the 236 mm x 127 μm i.d. emitter capillary inside the ESI probe assembly), it promotes the spiked peak profiles and spray pulsing. To further provide evidence for the contribution of post-decompression volume to37 spray pulsing, the restrictor capillary was detached from the ESI probe. Without any additional volume at the outlet of the restrictor, the observed decompression of the mobile phase was very smooth and

Alth of sp volum drople The p Suppo shape impro precis comp Intere reduce emitte peak s rate o reduce signal the m To restric isocra 1.0 m restric have p was re next i Val analyt human used a Since subseq six no curves Lin determ levels Suppo

Article

cutive pump green , and

er of not f the ictor. , the 2%), hout ke-up also erved

Figure 3. (A) spiking due tofrom inconstant pulsing thatof Figure 7: (A) Spiky Peaks or jagged peaks resulting inconstant spray spray pulsing because possible phaserandomly. separation. (B) Better peak shapes using the connector occurred (B) Observed peaks in prototype the absence of discussed spray in Figure 6. From ref [41]. pulsing.

Using the split-flow UHPSFC-MS interface, the portion of mobile phase routed toward the ESI source through the restrictor capillary is under direct influence of the mobile phase flow rate, mobile phase composition, and the BPR pressure 38 setting.41 Once the mobile phase flow rate, column temperature, and BPR pressure are set, only the viscosity of the mobile phase can affect the portion of the mobile phase directed to the MS through the restrictor capillary. The viscosity of the mobile

Figure 8: Solubility data of phenanthrene as a function of pressure at different temperatures in neat CO2 . Data obtained from ref [42].

39

2

different fixed backpressure values, 120 bar (A1 and B1), 150 bar (A2 and B2) and 180 bar (A3 and B3).

make-up pumps) and SFC mobile phase composition variations on the amount of MeOH collected per time unit. First, preliminary observations showed that an increase in the backpressure lead to a higher amount of collected MeOH. This effect is predictable because the higher resistance generated by the BPR obviously redirects a more important part of the mobile phase toward the transfer line. In the same way, an increase in the make-up pump flow rate or in the SFC mobile phase MeOH content evidently extended the measured amount of MeOH in the trap. In contrast, faster SFC mobile phase flow rates with constant MeOH composition led to a significant reduction in MeOH directed toward the transfer capillary. This counterintuitive effect is related to the fact that the active BPR leg of the splitter compensates pressure generated by the mobile phase to maintain a constant outlet backpressure. In the event of higher flows, the BPR is more open and offers a path of least resistance and hence the MS capillary receives less flow. Now, in contrast, when the mobile phase flow rate is low, the BPR must close in order to maintain system pressure, hence generating greater resistance along that leg and send more flow through the MS capillary.

Based on these preliminary observations and considering the tubing dimensions and Eq. (1), the MeOH flow rate entering the ESI probe was modeled over the whole range of chromatographic conditions using Excel. Some examples are shown in Fig. 4, including different mobile phase compositions (CO2 /MeOH: 95/5 (v/v) for (A) surfaces and 80/20 (v/v) for (B) surfaces) and various backpressures (120, 150 and 180 bar for surfaces 1, 2 and 3, respectively). As shown, the shape of these surface responses confirms the previous experimental observations. The validity of the model was then experimentally verified with 12 randomly selected sets of flow rates, backpressures and an extended set of mobile phase compositions (between 2% and 40% MeOH in CO2 (v/v)). Less than 10% difference was measured between the amount of MeOH physically collected in the trap and the value predicted by the computer simulation, which can therefore be considered as valid. In addition, it is worth mentioning that thanks to the dynamic adaptation of the split ratio between transfer capillary and the BPR depending on the chromatographic conditions, the total flow rate of MeOH entering the ESI probe tends to be leveled. The MeOH flow rate is always

Fig. 5. Model of the split ratio (A) and UHPSFC mobile phase dilution factor caused by the sheath flow (B) for the pre-BPR splitter with sheath pump interface as a function of UHPSFC mobile phase flow rate (x-axis) and sheath pump flow rate (y-axis) for a fixed UHPSFC mobile phase (CO2 /MeOH) composition of 90/10 (v/v) and 150 bar backpressure.

Figure 9: (A) Estimated values of split ratio and (B) dilution factor of the mobile-phase caused by the mobile-phase flowrate and the sheath (make-up) flowrate. Composition of the mobile-phase was kept fixed at 90/10 (v/v,%) of CO2 /methanol. ABPR was maintained at 150 bar. From ref [27]

40

Figure 10: Plots of isenthalpic or constant-enthalpy (in red) and isopycnic or constant-density (in blue) curves of pure CO2 on pressure-temperature plane. The values in red have kJ/g as the dimension and that in blue have g/mL. Data to generate these curves were obtained from REFPROP [36]. For more descriptions of the usage of these curves consult references [43, 47].This figure is used to explain the effect of connector heating/cooling in section 3.4.1.

41

Figure 11: Plots of isenthalpic or constant-enthalpy (in red) and isopycnic or constant-density (in blue) curves of 70/30 (mol/mol,%) CO2+methanol mixture on pressure-temperature plane. The values in red have kJ/g as the dimension and that in blue have g/mL. Data to generate these curves were obtained from REFPROP [36]. For more descriptions of the usage of such curves consult references [43, 47]. This figure is used to explain the effect of connector heating/cooling in section 3.4.1.

42

J. Duval et al. / J. Chromatogr. A 1509 (2017) 132–140

. 6. Influence of the flow rate of MeOH as a make-up solvent located after the BPR on the MS response of compounds. Analytical conditions: 2 Kinetex C18 colum ◦ Temperature: C, Outlet pressure (BPR): Injected volume: ␮L, SFC flow rate: APCI, Nebulizing gas: 0.04 MPa, Corona discha 0 × 4.6 mm, 2.7 ␮m),Figure 12: 9Demonstration of 10 theMPa, differences in 5variations of 1.2 MSmL/min, response as a function ◦ 000 nA, APCI temperature: 450 C, Dry gas: 3.5 L/min.

of make-up flowrates. For the hydroxylated triacylglycerols, MS response increased with increasing flowrates. For fatty-acids, on the other hand, the response decreased. For some analytes the response was not affected by flowrates. Consult ref. [21] for more details.

43

40

Y. Fujito et al. / J. Chromatogr. A 1508 (2017) 138–147

g. 3. The comparison between the normalized signal intensity of 441 compounds under three different flow rates of make-up solvent using the ratio of the peak area for 05, 0.1, and 0.4 mL/min to that of 0.2 mL/min in SFC/MS.between The ratio of normalized the peak area of signal each compound was plotted in order of elution. Figure 13: Comparison intensity of 441 compounds, conducted

with three different flow rates of make-up solvent - 0.05, 0.1, and 0.4 mL/min. Results show a column oven and a back pressure regulator (BPR) was the highest response was achieved with 0.1injector, mL/min flowrate. Consult ref [45] for more for SFC. Nexera UC SFC system employs low dead volume BPR. used details.

3. SFC conditionsthat

The Shimadzu Nexera UC supercritical fluid chromatography ystem equipped with two binary pumps, a CO2 pump, an auto-

This enables transfer of the total eluate from the analytical column to MS without diffusion of peaks, even when BPR and MS are con-

44

Chromatogr. A 1037 (2004) 393–403

Figure 14: Effects of changing temperature and composition for a two-component mixture. Starting from point (1), with a temperature of 75 o C and pressure of 150 bar, the method changes the temperature to 25 o C and then takes the composition to 75 %mol of methanol in CO2 . Depressurization process at this condition is shown by point (5). Consult ref [46] for more details.

45