Accelerating Prototyping Experiments for Traveling Wave Structures for Lossless Ion Manipulations

Abstract

Traveling wave structures for lossless ion manipulation (TW-SLIM) has proven a valuable tool for the separation and study of gas-phase ions. Unfortunately, many of the traditional components of TW-SLIM experiments manifest practical and financial barriers to the technique’s broad implementation. To this end, a series of technological innovations and methodologies are presented which enable for simplified SLIM experimentation and more rapid TW-SLIM prototyping. In addition to the use of multiple independent board sets that comprise the present SLIM system, we introduce a low-cost, multifunctional traveling wave generator to produce TW within the TW-SLIM. This square-wave producing unit proved effective in realizing TW-SLIM separations compared to traditional approaches. Maintaining a focus on lowering barriers to implementation, the present set of experiments explores the use of on-board injection (OBI) methods, which offer potential alternatives to ion funnel traps. These OBI techniques proved feasible and the ability of this simplified TW-SLIM platform to enhance ion accumulation was established. Further experimentation regarding ion accumulation revealed a complexity to ion accumulation within TW-SLIM that has yet to be expounded upon. Lastly, the ability of the presented TW-SLIM platform to store ions for extended periods (1 s) without significant loss (<10%) was demonstrated. The aforementioned experiments clearly establish the efficacy of a simplified TW-SLIM platform which promises to expand adoption and experimentation of the technique.

Graphical Abstract

Ion mobility spectrometry (IMS) is an analytical technique which separates gas-phase ions via their mobility, which is loosely related to physical size, as they traverse the medium while being driven by an electric field.^1,2 Renowned for their speed of analysis and relative simplicity, IMS techniques have seen widespread implementation in the fields of threat detection^3–5 and the investigation of physiologically relevant biomolecules.^6–8 More recent analyses have even leveraged IMS for the separation of isomeric species, something that has proved troublesome for many analytical techniques.^9–11

The most common implementation of IMS is drift tube IMS (DTIMS) which uses a linear, homogeneous electric field to measure gas-phase mobility coefficients. The ability of a DTIMS to separate ions of similar mobility is quantified as resolving power (Rp) which, when limited only by diffusion, is defined as:

$$R_p=\frac{t_D}{\Delta t}\approx \frac{1}{4}{\left(\frac{qLE}{k_{b}Tln\left(2\right)}\right)}^{1/2}$$ (1)

where t_D is an ion’s drift time, Δt is the peak width at half-maximum, q is the ion charge, L is the length of the drift tube, E is the electric field, k_b is Boltzmann’s constant, and T is the temperature.^2,12 Like many analytical techniques, the ability to separate increasingly similar species remains a focal point of experimentation and investigation in the field of IMS. According to equation 1, Rp can be improved by increasing the strength of the electric field, elongating the drift tube, or decreasing the temperature. While conceptually simple, these approaches to increasing resolving power suffer from practical limitations.^13–15 Given that implementing temperature controls requires more complex instrumentation and robust electronics, lowering the temperature of the system is not always feasible nor practical for many experimental scenarios.¹⁶ While increasing the field strength could, in theory, simply be accomplished by “turning a knob” it is important to note that higher fields require highly specialized power supplies and can contribute to electrical breakdown. Likewise, given the linear nature of the fields within a DTIMS, increasing the length of the system would require increasing the voltage at the start of the drift region to maintain the desired electric field, once again requiring capable power supplies and contributing to electrical breakdown.

While certainly the most recognizable, the DTIMS is not the only IMS technique in use today. Debuting in the early 2000s,¹⁷ traveling wave IMS (TWIMS) has seen commercial implementation and widespread application.^18–20 Unlike DTIMS which applies a static linear electric field, the electric fields in a TWIMS are transient and dynamic.^21,22 In TWIMS, separation is achieved as the frequency at which ions “roll over” waves is dependent upon their mobilities.²² The more frequently an ion rolls over, the greater its arrival time at the detector. Perhaps one of the biggest advantages of TWIMS is that the length of the separation region can theoretically be increased without increasing the voltage at the start of the region. This is due to the transient nature of the waves which leads to local minima and maxima, as opposed to absolute minima and maxima required for DTIMS.

While promising, there remain several hurdles surrounding the advancement of TWIMS. For instance, while the length of a TWIMS is not limited by issues of electrical breakdown, there remains practical limitations on length with regards to the physical size of the instrument. These limitations are of particular concern for commercial instrumentation which must be assembled and housed in a fashion that makes for reasonable placement/storage in a laboratory. Additionally, the electrode structure of commercially available TWIMS limit the range of traveling wave and RF amplitudes that can be applied during analysis, which ultimately constrains the range of m/z values that can be simultaneously analyzed with the system.²³ Behaving in a fashion similar to a stacked ring ion guide, there is a window of m/z and mobility values that allow for stable ion motion within the TW-SLIM system.^24,25

A potential solution to the shortcomings of traditional TWIMS, known as traveling wave structures for lossless ion manipulation (TW-SLIM), debuted in 2015 and has already seen a wide range of applications and modifications.^26–30 Comprised of inexpensive and readily customizable printed circuit boards (PCB), TW-SLIM offers a simplified electrode structures which, when paired with the low cost of PCBs, lends to rapid and creative prototyping. One of the most significant advancements in TW-SLIM has been the implementation of serpentine ultralong paths with extended routing (SUPER) which enable for theoretically infinite pathlengths for separation.³¹ It is worth noting that while other TWAVE instruments such as the cyclic ion mobility system developed by commercial vendors (e.g. Waters) attempt to address the issues of limited pathlength, the planar geometry of TW-SLIM is more amenable to generation of extended pathlengths.³² Multiple laps do, in fact, mitigate the issues associated with limited path length but also present a unique set of additional challenges related to issues of ion population lapping which can confound spectral interpretation. Nevertheless, with greatly expanded pathlengths and simplified electrode structures (See Figure S1) that limit traditional TWIMS techniques, the possibility for TW-SLIM to catalyze the advancement of TWIMS and IMS as a whole, is apparent.

Despite showing great promise, there remains a series of obstacles that stand between TW-SLIM and its ultimate potential as a high-powered, low-cost gas phase separation technique. While SLIM may be constructed of inexpensive materials, the technology used to generate the waveforms in the SLIM requires a degree of sophistication which manifests both applicative and financial barriers to its implementation. Additionally, more simplistic TW-SLIM experiments typically rely on ion funnel traps (IFT) for injection which require levels of fabrication precision that are not readily available to most laboratories. Lastly, the primary means of detection in TW-SLIM experiments are mass spectrometers (MS). While MS is undoubtedly a powerful analytical tool, the issue of cost as well as compliance with pressure requirements can limit implementation and slow prototyping efforts.

Recognizing the obstacles to TW-SLIM implementation, the following manuscript outlines an instrumental platform capable of performing the basic operations typical of a TW-SLIM, while utilizing an expansive array of techniques which reduce the practical, financial, and intellectual barriers to experimentation. With regards to TW generation, the efficacy of a simple and inexpensive multifunctional traveling wave generator (MFT) was evaluated and was found to perform comparably to more traditional approaches. Concerning ion injection, an on-board injection mechanism (OBI) was employed which mitigated the need for an IFT, simplifying the experiment. The ability of this OBI to simultaneously aid in on-board ion accumulation (OBA), as outlined elsewhere,^33,34 was evaluated as well. Regarding ion detection, this report details the use of an appropriately shielded Faraday plate designed specifically for the TW-SLIM experiment serving as an inexpensive alternative to the MS techniques or highly complex charge detection schemes previously reported.³⁵ Lastly, the potential for this simplified TW-SLIM platform to aid in more complex experimentation was evaluated as it proved capable of storing ions for extended periods of time (1 s) with minimal loss.

Experimental Section

Tetraoctyl ammonium bromide (T8, nominal mass: 466 Da) and tetradodecyl ammonium bromide (T12, nominal mass: 690 Da) were purchased from Sigma-Aldrich (St. Louis, MO). Morphine (nominal mass: 285 Da) standards were purchased from Cerilliant (Round Rock, TX). All data presented were collected using a solution of 10 μM morphine and 1 μM each of T8 and T12 prepared in methanol (HPLC grade, Fisher Chemical, Fair Lawn, NJ) with 0.1% formic acid (Honeywell-Fluka, Muskegon, MI).

A schematic of the instrument used throughout the presented work is shown in Figure 1A. Ions were generated via electrospray ionization (ESI) using a 75 μm ID glass capillary emitter with a sample flow rate of 3 μL/min, controlled by a Model 11 Syringe Pump from Harvard Apparatus (Holliston, MA). The ESI emitter was biased 2.2 kV relative to an adjacent stainless-steel inlet capillary (560 μm i.d., 1600 μm o.d.) purchased from McMaster-Carr (Elmhurst, IL). Having traversed the heated (160 °C) capillary, ions were focused by an ion funnel (897 kHz, 240 V_p-p) housed in a low pressure (2.45 torr) chamber before being transported to a second, smaller ion funnel (1.1 MHz, 140 V_p-p) housed in a slightly higher pressure (2.50 torr) chamber. The gas inside each of the chambers was ambient air as the focus was on the characterization of the drive electronics and not theoretical ion-neutral cross section accuracy. Pressure was measured in each chamber using Setra Model 730 capacitance manometers (Boxborough, MA). The disparate chamber pressures and offset of the capillary as depicted in Figure 1A were implemented to limit the transmission of neutral solvent molecules into the second chamber. After traveling through the second ion funnel, ions arrived at the first TW-SLIM segment referred to as the “drag” board. Electrodes designated as blocks A, B, and C (Figure 1B) on the drag board could be operated either as TW electrodes or as blocking electrodes, in which case they were biased above the adjacent electrodes using a static DC potential. In the context of the presented work, block B was used as a point of ion packet injection, serving the role of an ion gate found in traditional ion mobility experiments. Upon switching from the block to the TW state, block B would permit the transmission of ions. The width of the injection is defined as the period in which block B operated as TW electrodes. The use of block A as either TW or blocking electrodes was varied between experiments and will be discussed in further detail when appropriate. For all the experiments discussed, block C remained in the TW state.

Figure 1.

(A) Schematic of the instrument used throughout the presented work. (B) Annotated images of the “drag” board and (C) “zipper” board. Images are not to scale.

Having navigated the drag board, ions then reached the interface between the drag board and a second TW-SLIM segment, referred to as the zipper board in reference to the track design (Figure 1C). The zipper design was implemented as it provides enhanced pathlenghts for separation without greatly increasing the length of the PCB. Unlike previous iterations of the TW-SLIM experiment, this setup uses two distinct sets of SLIM segments, located adjacent to each other, as opposed to a singular set of SLIM boards. The impact of this multi-segmented approach and the flexibility associated with its implementation will be explored in future publications. The total pathlength used for separation from the point of injection to the end of the zipper was approximately 70 cm. A custom Faraday plate was placed directly adjacent to the end of the zipper board and served as the primary detector. A detailed schematic of the custom Faraday plate is provided Figure S1. Current detected at the Faraday plate was amplified using a Keithley Model 427 current amplifier (10⁹ Gain, Rise time: 1 ms, Cleveland, OH) and the signal was recorded as a function of time using an Analog Discovery 2 (Digilent Inc., U.S). All data presented in this manuscript were derived from spectra averaged 100 times. Each data point represents an average of 5 spectral replicates with the standard deviation of those 5 replicates representative of the uncertainty of the measurement. Peak identities were assigned by analyzing individual standards of each of the 3 analytes (morphine, T8, and T12). Data were processed using Python notebooks developed in-house. A sample notebook is provided in the supporting information.

The following TW-SLIM conditions were used throughout all the presented experiments and were controlled using a modular intelligent power source (MIPS) from GAA Custom Electronics (Kennewick, WA) unless otherwise stated. Ions were confined laterally by guard electrodes biased 11 V to the TW, and radially by RF electrodes operating at 1.27 MHz, 300 V_p-p. The TW amplitude was 22 V_p-p and the TW speed was 128 m/s. Further details regarding the configuration of the SLIM boards can be found in the supporting information. To be clear, the outlined TW conditions were selected for optimized ion transmission and storage experiments and were not optimized for resolving power. While conditions which reduce peak width and subsequently enhance resolving power are achievable, it is worth noting that observed peak widths did not differ substantially from those predicted by theory for a single pass TW-SLIM experiment and the minimal path length used (Figure S2).¹²

Compared to the MIPS system, which uses 8 different DC channels to generate the associated TW, the MFT uses a single DC supply to generate all of the TW signals. For experiments involving the MFT, a SPD3303X-E programmable DC power supply purchased from Siglent (Solon, OH) was used to provide the DC bias which was modulated by the MFT to produce the TW. Key to the MFT functionality is the multichannel switch that is commonly used for medical ultrasound applications (e.g., Maxim 14803). Interestingly, the electrical requirements for ultrasound applications and the modulation of multiple electrodes are directly analogous to the TW experiments. Additionally, it is worth noting that the MFT in its current configuration is controlled by a M0 microcontroller (Adafruit Trinket M0) and contains the same core functionality as the MIPS including TW generation, frequency adjustment, forward and backward TW movement, float capacity, and TW halt functionality. The system additionally accommodates an external trigger to cycle through pre-programmed events.

Results and Discussion

In order to evaluate the efficacy of the MFT as a TW generator, its performance was placed in the context of its more traditional counterpart, the MIPS. This was accomplished by operating the TW-SLIM in two modes: “MIPS mode” where both the drag and zipper TW were generated by the MIPS, and “MFT mode” where the drag TW was generated by the MIPS and the TW of the zipper was provided by the MFT. Mobility spectra collected when operating the instrument in these different modes provided the foundation for analyses of the MFT performance. While preliminary data indicates that both the drag and zipper TW can be effectively generated by a single MFT, an emphasis was placed on evaluating the ability of the MFT to work in tandem with a MIPS in the presented analyses. For all comparisons between the MIPS and MFT mode, block A remained in the TW state throughout the analyses and no TW conditions that favor accumulation were implemented. Initial testing revealed an apparent variation between spectra collected in the MIPS and MFT modes, as there was a noticeable shift to later arrival times and subsequent increases in peak width when operating in MFT mode (Figure 2A). This shift is attributed to subtle but real deviations (~ 0.3 V) from the desired TW amplitude present in each system, which when compounded, resulted in the behavior shown in Figure 2A. It is worth noting that if analyses hinge on tightly aligned TW generated within a system in which MFT and MIPS work in tandem, it is possible to tune the TW amplitude of either the MIPS or MFT so that the system performs similarly to how it would with just MIPS or MFT (Figure 2B). For clarity, the spectra shown in Figure 2B correspond to actual spectra where the MFT TW amplitude was tuned to match that of the MIPS, and not alignment through digital signal processing.

Figure 2.

Mobility spectra collected when TW were generated in MIPS mode (blue) and MFT mode (red). The spectra on the top (a) are misaligned due to variations in TW amplitude which, once corrected for, produce the spectra on the bottom (b). The injection width used when collecting the presented spectra was 1 ms, with the exception of the spectra labeled “MIPS (200 μs Inj. Width).” Peaks at (approximately) 10, 27, and 45 ms correspond to morphine, T8, and T12, respectively.

A more thorough analysis of the MFT’s performance was accomplished by comparing numerous spectral metrics for each analyte, across different injection widths, when operating in MFT mode relative to MIPS mode (Figure 3). For these analyses, the amplitude of the TW generated by the MFT was not corrected to align with that corresponding MIPS spectra. Consequently, as observed in Figure 2, the arrival times are consistently greater when operating in MFT mode relative to MIPS mode. This increased arrival time naturally coincides with the observed attenuation in peak amplitude as peaks broaden due to ion populations undergoing diffusion for a longer period. It is important to note that while peak amplitudes were slightly diminished when operating in MFT mode, peak areas remained comparable, with no peak area varying more than 10% relative to its respective value generated when operating in MIPS mode. This suggests that the diminished peak amplitudes are in fact the result of diffusion and not ion loss. The most substantial difference between the two modes is observed for T12 when operating with a 100 μs injection width. This pronounced deviation is attributed to an observed high mobility bias of the injection mechanism which limited the population of the injected T12 ion packet (Fig 2b). This type of behavior is well within expected levels of performance as mobility bias is well documented for modulated ion populations.^36–38 Unsurprisingly, the diminished ion population exhibited greater run-to-run variability which manifested as greater deviations from the corresponding MIPS mode values and greater deviations between measurements. Considering that the deviations observed in Figure 3 can be largely attributed to the TW amplitude generated by the MFT differing from that of the MIPS (which can be mitigated relatively easily) and the presence of a mobility bias in an injection method, the comparable performance of the MFT mode relative to that of the MIPS is extremely promising. Given the relative simplicity and low cost of the MFT, its ability to operate in-tandem with a MIPS expands the suite of simultaneous TW-SLIM experiments to augment future TW-SLIM configurations and increasingly complex TW-SLIM designs.

Figure 3.

Spectral metrics for each analyte, across different injection widths, when operating in MFT mode relative to MIPS mode. An agreement of 100% would indicate that the value attributed to MFT mode is identical to that which was collected when operating in MIPS mode. Raw data associated with these plots can be found in Table S1.

Much like the MFT, the use of OBA and OBI offers a simple and serviceable alternative to a more traditionally used technique: the ion funnel trap.^25,39 While used in the context of ion injection in the MFT evaluation, static potentials have been used in other TW-SLIM experiments which exploit the nature of TW to allow for the accumulation of high density ion populations relative to the continuous current generated by an ESI emitter.^34,40,41 Given the apparent benefits to sensitivity that come with ion accumulation, experiments were performed to characterize the ability of the presented instrument to accumulate large ion populations. For simplicity, the MFT was excluded from these analyses as the emphasis was on ion accumulation, not TW generation. Prior to injection, the TW operates under conditions aimed at increasing the likelihood of ion roll over (lower TW amplitude and or higher TW speed) akin to the approach outlined by Deng et al. which has shown to increase ion accumulation in TW-SLIM.³⁴ Per the nomenclature established by Deng et al., these conditions are referred to as a “gentle” traveling wave (GW).³⁴ In our implementation, after some period defined as “accumulation time,” block A is raised to a static DC bias, block B transitions to a typical TW electrode for some period (i.e., injection width), and the TW returns to its usual conditions used for separation (22 V_p-p, 128 m/s). These separation conditions are maintained until the SLIM has been cleared of all injected ions after which point the timing cycle repeats. The transition of block A to a static DC bias during the injection period was done to ensure that the ion population being injected consisted only of ions accumulated when the GW was implemented.

The first set of experiments regarding accumulation were aimed at evaluating how different GW conditions impact the accumulation of different ion species. GW conditions in which TW amplitude was held constant while TW speed was increased, and conditions in which TW speed was held constant while TW amplitude was lowered, were applied. Variations in the observed peak areas for morphine and T12 as a function of different GW conditions are presented in Figure 4. The accumulation time used for these experiments was 15 ms and the injection period was 3.2 ms. The 3.2 ms injection was chosen as it provided enough time across all conditions to allow for the region between blocks A and B to be cleared of ions. Data recorded at 22 V_p-p and 128 m/s represent a baseline in which no GW was applied and ions were accumulated under normal TW conditions. When achieving GW conditions using increased TW speed, both morphine and T12 show an increase in peak area relative to their baseline values. Interestingly, morphine observed greater relative gains to its peak area using the reduced TW speed approach compared to T12 which showed modest and, in some cases, negligible gains. In agreement with results published by Deng et al., decreasing TW amplitude proved to have the greater impact on the accumulated ion populations compared to varying TW speed.³⁴ Peak areas for both compounds were consistently greater when using the reduced amplitude GW, except for that of T12 when a TW amplitude of 12 V_p-p was applied. The noticeable depreciation of the T12 peak area as the TW amplitude decreased from 18 to 12 V_p-p is attributed to poor confinement of T12 ions. In contrast, morphine shows only a minor decline in peak area and only at the lowest TW amplitude evaluated (12 V_p-p). These distinctly different trends in peak area for morphine and T12, suggests that considerations should be made regarding factors that influence confinement (guard bias and RF amplitude/frequency) as well as the range of mobilities being analyzed when accumulating ions in TW-SLIM.

Figure 4.

Plot of peak area vs TW amplitude and speed. Data collected when the TW amplitude was held constant (22 V_p-p) while TW speed was increased are shown in red. The data shown in blue were collected when the TW speed was held constant (128 m/s) and the TW amplitude was reduced.

Having established that the presented instrument could effectively accumulate ions, an additional round of experiments was performed in which the impact of accumulation time on the population of differing ion species was evaluated. The same experimental workflow for accumulation outlined previously was used for these analyses as well; however, the GW conditions used this time were static, 16 V_p-p and 128 m/s, and the injection width was 2 ms. The GW conditions were selected based on the results shown in Figure 4. An injection width of 2 ms was chosen as it had proven capable of providing ample time to clear all ions from the region between blocks A and B for the GW conditions chosen for these analyses. The impact of accumulation time on ion populations is presented in Figure 5A. Similar to observations made by Deng et al., the total ion current (approximated here as the sum of the peak areas of morphine, T8, and T12) increases as a function of time until eventually plateauing.³⁴ Interestingly, plots of peak area vs accumulation time for each compound individually reveal behavior that deviates from that of the total ion population. With exception of the largest species, T12, ion populations reach a local maximum at a given accumulation time (10 ms for morphine, 30 ms for T8), after which point the ion population begins to decrease before eventually stabilizing. Figure 5B displays peak areas for each of the 3 compounds, normalized to their apex, as a function of accumulation time. Both the accumulation time at which the local maxima is reached for each compound and the magnitude of the decrease that follows appears to be related to the size of the ion species. Morphine, being the smallest of the ion populations, reaches its maxima before T8 and experiences a decline of 49±2% relative to its apex, a decline far greater than that of T8 (16±2%). The apex for T12, the largest species, coincides with the plateau of the total ion current and shows no significant decline. While limited in scope, these results suggest that as ions are allowed to accumulate for longer and longer periods, space charge effects begin to play a considerable factor in determining which ion species dominate the accumulation space, with higher abundance species “pushing out” those with smaller contributions to the total charge in the system. While not surprising given reports and personal experience with ion funnel traps,^42,43 these results remain notable when considering OBA and OBI approaches for TW-SLIM experiments. Similar to the results of the GW condition experiments, this evaluation of the impact of accumulation time suggests that when accumulating species of disparate size in a TW-SLIM, it is imperative to recognize that a “one size fits all” approach may be detrimental to select ion populations. Recent reports by commercial entities regarding OBA and OBI using TW-SLIM also note potential challenges related with OBI and OBI but the data shown below provide a quantitative accounting of this behavior.⁴⁴

Figure 5.

(a) Plot of Peak area vs accumulation time for the 3 analytes individually, as well as the sum of the three (“Total”). For this experiment “Total” is used to approximate the total ion current, given that the three analytes are the only ones present in the solution. (b) Plot of peak area for each of the three compounds normalized to their greatest value as a function of accumulation time.

The results of the experiments discussed previously suggest that the presented instrument can perform many of the basic operations typical of a TW-SLIM experiment. As a final assessment of the presented TW-SLIM platform, the potential for this set-up to be used for more complex TW-SLIM experiments was evaluated. More specifically, the ability of this instrument to store ions for extended periods of times, a technique that could enable a user to probe issues of gas-phase chemistry, was evaluated. For these experiments, ion storage was accomplished by first allowing ions to enter the TW-SLIM and then subsequently halting the TW entirely, creating a series of static ion traps throughout the SLIM. Two different approaches to ion storage were evaluated: a post-injection halt in which ions were injected, allowed to separate for 5 ms, and then halted, as well a pre-injection halt in which the TW was halted while ions were trapped between blocks A and B prior to injection. Normal TW conditions were resumed after the halt period. The conditions for these experiments were identical to those of the accumulation time evaluation but with a static accumulation time of 15 ms and the added halt step. For both the pre- and post-injection experiments, the TW amplitude was 22 Vp-p during the halt period.

During the initial evaluation of ion storage within TW-SLIM, the MIPS was used to halt and store ions within the system. These experiments served as the baseline for comparison to the MFT. With the capacity to accept an external trigger to adjust TW parameters, the MFT can directly mimic the functionality of the standardized MIPS platform. For the sake of brevity, the full complement of storage experiments was not repeated as the capacity to match the MIPS functionality with the MFT was effectively demonstrated in Figures 2 and 3. Storage experiments conducted using the MFT demonstrated the capacity to control the TW modulations required for extended ion storage. Figure S3 illustrates the recorded ion signal using the MFT for a condition where no halting was observed and an ion population held within the SLIM system for 1 s using a post-injection halt.

Results showing the impact of storage time on the observed area and intensity of the T12 peak are shown in Figure 6. For both peak area and amplitude, the reported values were normalized to their respective values when no storage was implemented. Sample spectra collected with no halt and a 1 s halt period are shown in Figure S4. After halting the ions for a period of 1 s, the post-injection data for both peak area and amplitude depreciates less than 10% relative to the unhalted data, whereas the pre-injection data shows losses of approximately 50%. It’s also worth noting that both peak areas and amplitudes from the post-injection halt experiment remained relatively static, whereas the pre-injection data decreases as a function of accumulation time until 50 ms, at which point the values begin to plateau. Considering that the pre-injection experiment does not allow for ions to separate prior to storing them, the most likely cause of the significant loss and the trend of said loss is space charge effects, as a large population of ions is confined between blocks A and B. When ions are allowed to spread out on the board (i.e., the post-injection halt) these detrimental effects are not observed. While some loss of ions appears inevitable when storing with the presented approach, allowing ion populations to disperse prior to storing them mitigates the losses, indicating that the proposed platform could potentially serve as a tool for the exploration of gas phase ion chemistry.

Figure 6.

Variation in peak area (top) and amplitude (bottom) as a function of halt time. “Post-injection halt” refers to data collected when the ions were stored post injection, after 5 ms of separation. “Pre-injection halt” data was collected when ions were stored in the region between blocks A and B, prior to injection.

Conclusions

The presented multi-board TW-SLIM platform realizes the fundamental operations typically performed in single-board TW-SLIM experiments, while offering several simplifications that lower the barriers to its implementation. Recognizing that the MFT as a TW generator is considerably less complex and more economical than its traditional counterpart, the MIPS, its impressive performance implies it may play a role in augmenting future TW-SLIM experiments. By using MFTs, one could perform TW-SLIM experiments in which numerous distinct TW are generated within a single TW-SLIM, allowing for more complex experiments with a wide range of functionality. Like the MFT, the use of OBA and OBI simplifies the TW-SLIM experiment, removing the need for an ion-gating mechanism independent of the TW-SLIM. While the nature of the OBA and OBI does produce a mobility bias, optimization of the TW conditions during the injection period could potentially allow a user to fine-tune the range of mobilities effectively injected by the OBI. While not explored in the context of this manuscript, the use of distinct SLIM segments does not appear to adversely impact the TW-SLIM experiment. A more thorough investigation on the impact of distinct SLIM segments will be presented in a future publication. The use of multiple distinct SLIM segments could allow a user to construct a diverse array of TW-SLIM configurations using a series of interchangeable SLIM boards of varying functionality. Lastly, to our knowledge, the presented TW-SLIM platform is the first to use a shielded Faraday plate as a detector to collect raw spectra. More sophisticated charge detection mechanisms have been reported previously but have yet to demonstrate spectra with comparable signal to noise ratios.³⁵ While limited in its functionality, the Faraday plate is far more economical than other detectors and eliminates concern regarding the process of venting a mass analyzer. The ability to rapidly bring the instrument to and from vacuum allows for more frequent adjustments to this system which affords considerable benefits when prototyping new TW-SLIM designs.

Though the scope of the experiments conducted with this simplified TW-SLIM platform is limited, the value of such a system has been shown already as it has enhanced our understanding of the underlying mechanics of SLIM accumulation. While a simple approach to accumulation may degrade the quality of analysis for some species, understanding how accumulation conditions affect given ion species may allow a user to selectively improve the abundance of a given species of interest. Furthermore, this TW-SLIM platform has shown potential as a tool for the exploration of ion chemistry as it is capable of storing ions for extended periods (1 s), allowing a user ample time to introduce reactants into the region where these ion populations are stored. The implementation of storage times that exceed 1 s were precluded from the presented work but will be explored in future publications as the trends shown in Figure 6 suggest greater storage times are achievable. The development of TW-SLIM platforms such as the one presented in this manuscript are critical as they allow for more widespread implementation of the technique by introducing a modular assembly approach and reducing costs while increasing flexibility using the MFT.

Manuscript Highlights.

• Simplifying the means of traveling wave generation lowers prototyping barriers

• Open-sourced electronic design facilitates broad adoption and experimentation

• Multi-board SLIM configuration introduces design and layout flexibility

• Judicious choice of onboard ion accumulation conditions impacts trapping efficiency