Patent Application
20090076294
This application relates to a method of controllably inducing nucleation of a solute in a solution having a net charge to promote the growth of crystals.
Crystallization of solutes from solution is an important industrial separation and purification process. The first step in the process is the creation of the new phase which is known as nucleation. Crystallization is thus preceded by nucleation, which occurs either spontaneously or is induced by particles or vibration. Experimental studies of the thermodynamics and kinetics of nucleation processes are difficult because of the role of surfaces and impurities in aiding the nucleation process.
The formation of crystals of a solute (i.e. a dissolved solid) is typically described in terms of solute's phase diagram in a given liquid medium. The medium may be defined by its chemical properties (such as concentration of and types of solvents, electrolyte, pH, buffers, impurities, and other solute(s) of interest) and physical properties (such as type of container, temperature, pressure, magnetic fields, electric fields, and gravity). Provided the phase diagram is known, or can be established through experiment, crystals of a dissolved solid can in principle, be obtained. For example, in the case of crystallization of inorganic compounds for which phase transitions and phase diagrams can be measured, theories of crystal nucleation and growth have been applied successfully.[1-7]
However, despite the development of solute nucleation and crystallization theory over a period of at least 100 hundred years, it is often difficult to demonstrate in practice that the formation of a macroscopic crystal of dimensions >1 μm requires the formation of nanoscopic nuclei. It is believed that nuclei on the scale of only a few nanometers form and disappear rapidly in the nucleation medium (i.e. in an equilibrium process). A key step is to coax, through adjustment of the chemical and/or physical description of the nucleation medium, those nuclei to grow to a critical size such that crystallization is induced. The degree of supersaturation of the solute is an important factor, and the solute concentration needed to cause nuclei to form and then grow into crystals is referred to as “critical supersaturation”. Once the nuclei has formed, that nuclei must reach a critical size that is dependent on each solute, such that the solute will spontaneously precipitate onto the nuclei causing it to grow into a crystal rather than shrinking in size and ultimately disappearing.
Theories of crystal growth do not necessarily help the experimentalist determine which conditions in the nucleation medium are necessary for nucleation and growth of crystals for compounds that have not previously been crystallized by experiment. Examples of such compounds include a newly isolated natural product or novel synthesized compounds (such as a new pharmaceutical compound or an analogue of it), or known compounds that have not yet yielded to crystallization attempts (such as soluble or integral-membrane proteins). The crystallization of novel pharmaceutical compounds is in fact regarded by those who synthesize them as an empirical process where proficiency is achieved through experience and trial and error.
These practical issues for solute crystallization have led to a large number of crystallization strategies that have resulted in an even larger number of experimental techniques and methods. Numerous investigators have studied organic compound nucleation and crystal growth in container-less (or wall-less) sample vessels levitated in a medium (typically air at atmospheric pressure). Wall-less sample preparation has been described in detail previously be the inventors [8]. The rationale for the use of container-less vessels in which to effect nucleation and crystal growth was to provide a more homogeneous environment in which a crystal could be grown because there was no liquid:solid interface at the wall of the nucleation vessel or medium (.ie. such as a plastic or glass vial).
One of the first, and most celebrated, experiments based on single particles was the Millikan oil drop experiment, performed in 1909.[9,10] By partially suspending single charged oil droplets in a DC field, Millikan was able to determine the charge of an electron, an experiment for which he was awarded the Nobel prize. Later modification of the experimental apparatus by Wolfgang Paul to include AC fields to trap ions under vacuum led to the development of modem quadrupolar field-based mass spectrometers, and another Nobel prize.[1] The Paul trap technology has since been extended to the trapping of more massive particles by suitable adjustment of the frequencies and amplitudes of the potentials applied to the electrodes of these devices. This resulted in the development of the electrodynamic balance for single aerosol particle characterization. This latter technology has been widely applied to the study of the physical characteristics of single levitated droplets/particles, such as evaporation, charge, and condensation processes.[12-14] Efforts to probe the chemistry occurring in these levitated particles has lead to elegant laser-based probing methods revealing real-time chemical reaction information relevant to atmospheric processes.[15]
An application of the container-less vessel (i.e. a levitated droplet) for nucleation of organic compounds which has received considerable attention is for protein crystallization. Again, the absence of an interface at the solution-container wall was believed to be of significance with respect to avoiding the growth of impure crystals or crystal with defects, and particularly so when the nucleation experiment was performed at low temperature. The types of levitation of containerless nucleation media (otherwise known as levitated droplets) are acoustic, magnetic, electrostatic, optical, and combinations thereof. Most of these types of levitation have been studied with respect to their utility in forming crystals of organic compounds.
A tremendous quantity of research has been performed on gas-phase ion-molecule reactions. In some cases solute nucleation and crystal growth in the presence of a charged entity has been performed. This body of work has clearly shown that ions can act as nuclei for the clustering and growth of aggregates. The structure of these clusters could be of relevance with respect to the nanoscopic nuclei that ultimately grow into crystals, and moreover, gas-phase ion-molecule reactions are widely acknowledged as experiments that provide snapshots of ion solvation in the (liquid) condensed phase. Though in these studies the aggregates are themselves isolated in the gas phase, the aggregate itself could or could not be in a solid or liquid form depending on the temperature of the aggregate.
The scope of ion-molecule reactions is enormous, and those studies span atomic ions and molecular ions as the charge center onto which molecules and atoms cluster around.[1,6] This work has now extended into the gas phase acid-base properties of proteins.[1,7] Synthetic diamond crystal growth is described as the formation of a 13-Carbon center using chemical vapor deposition technology to form C− in the gas phase that then causes nucleation of carbon as diamond, rather than the more thermodynamically stable graphite form.[18, 19]
Prior art studies have shown that it is often difficult to achieve stable nuclei in the gas phase (i.e. without decomposition). In practice it is difficult to achieve a high enough density in the gas phase to form nuclei that can actually give rise to a crystal. While some ion-molecule clustering and nucleation has been observed in the gas phase, analogous experiments which involve ions clustering phenomena in the condensed phase of a medium that has net charge have not been performed. A condensed phase medium that has net charge may mimic the net charge on a gas phase ion. However, condensed phase media in which the mass-to-charge ratio of that medium was adjusted has not been demonstrated or reported in the literature as an experimental variable that influences nucleation in the condensed phase. This is significant because in essentially all condensed phase media in which nucleation has been studied, regardless of the abundance of charged species present in such media, those media had an overall net electrical charge of zero (i.e. the nucleation media were neutral).
The possibility that the net charge (and thus the mass-to-charge ratio) of the nucleation medium could influence condensed phase nucleation or crystal growth does not appear to have been previously disclosed or demonstrated in the prior art. Tang et al. have described an investigation of solute nucleation in levitated solution droplets[20] This manuscript dealt with the determination of critical supersaturation for the nucleation of NaCl and NH4SO4 systems, which is a process believed to be important in tropospheric aerosols. However, there is no mention of variation of the net charge of the nucleation medium, nor the mass-to-charge ratio of the nucleation medium, which was a levitated droplet. As an another example of the literature from the atmospheric community regarding the transition from liquid phase to solid phase in aerosols, the Leisner group has published several reports on the nucleation of atmospherically relevant compounds in levitated droplets.[21,22] Still others use the electrodynamic balance to study reactions at the particle-air interface (ie. heterogeneous reactions),[23-26] and phase transfer and freezing processes.[27, 21, 28, 29] The motivation for conducting many of these studies was the hypothesis that reactions at the droplet/particle-air interface are of primary relevance with respect to understanding the science of the troposphere and stratosphere.
Other reports of chemistry in levitated droplets range from simple acid-base reaction in a picoliter vessel,[30] to reactions that were photochemically initiated.[31-35] Optical levitation has also been used to study similar reactions. Cederfelt and co-workers described the charge limit for a droplet in which NaCl could be crystallized in a levitated droplet.[36] Basically, they were interested in determining the maximum net charge that could be contained in a droplet which dried to a solid residue of NaCl without the droplet undergoing a Coulomb explosion. Coulomb explosion is a process by which a droplet with net charge fragments because the repulsive force of its net charge exceeds the attractive force of the droplet, which is a function of the droplet's surface tension and radius.[37-39] There was no mention in the Cederfelt and co-worker manuscript of a change in the nucleation rate or abundance of nucleation sites as a function of net charge (likely because they were nucleating NaCl which is arguably the easiest compound to form crystals of).
In related studies that are of direct relevance to Coulomb explosion, the charge loss from single levitated droplets has been described in a number of manuscripts in recent years.[38-45] The process of a droplet with net charge releasing some of that charge (i.e. Coulomb explosion) has received considerable attention because of the introduction of Electrospray Mass Spectrometry for the characterization of biomacromolecules by John Fenn.[46-48] Interestingly, a charged cluster, which is a collection of molecules together with one or more ions that is viewed as being an entity intermediate between the gas and condensed phases, can be produced in abundance in an Electrospray.[49-70] The propensity to produce charged clusters in an Electrospray has been recently exploited by Fernandez de la Mora in his experiments that involved studies of neutral molecule clustering around small charged clusters.[71] Again, this is reminiscent of the classic gas-phase ion molecule clustering studies discussed above. Protein clusters produced by an Electrospray have also been observed in the gas phase [73-76] and also after their deposition onto a surface.[77]
Many studies have used acoustic levitation to suspend droplets in air in which protein crystallization was studied.[78-80] One of the investigators, Staffan Nilsson, has claimed that his technique is of general utility because his methodology allows for experiments to be designed to learn of the optimal conditions in which nucleation is initiated while, importantly, each one of his experiments consumes only picoliters of sample solution.[81-83] A company in San Diego has also described a similar approach, with respect to the consumption of only picoliters of sample solution per experiment.[84] This company did not utilize levitation, but they were using similar droplet generators (ie. ink-jet style droplet-on-demand generators) and that was the foundation for their automation of the experimentation work needed to define optimal conditions for protein crystallization.[84]
Other groups have used a combination of acoustic and electrostatic forces to levitate a droplet, and in some reports, deliberately caused the levitated droplet to rotate slowly in an effort to simulate a space environment (i.e. the condition of microgravity) while proteins were allowed to crystallize.[85-91] Chung and co-workers have described at length how the net elementary charge in their levitated droplets was restricted to the surface of the levitated droplet.[86-87] To them, this was an important feature of their nucleation medium because they felt that a “homogeneous interior” volume of the droplet, and specifically that its net neutral charge was an important factor in allowing proteins to nucleate, and then grow into crystals. These authors described the volume of liquid surrounding the “homogeneous interior” as containing the net charge in the droplet. These authors gave the impression that the electric field on the surface of the levitated droplet (due to the net excess charge carried by the droplet) would (or could?) interfere with protein crystallization. However, it appears that the mass-to-charge ratio of the droplets levitated using a combination of acoustic and electric forces in the studies by Chung and co-workers were, well above the threshold necessary to observe effects on the nucleation of a solute as described below.
Another set of groups have been using optical trapping and its forces to promote crystal growth, and they can also transfer those crystals to a growth solution (ie. seeding).[92] Other groups are simply using the intense electric fields of a focused laser beam to induce nucleation.[93-96] There is one report of the use of laser ablation for crystal growth, and that report was applied to diamond growth.[97]
Of possible relevance to the crystallization of proteins are the types of interactions between proteins and various substrates. Theories of protein crystal growth are now emerging in which a protein, charged because its functional groups are protonated or deprotonated (ie. —NH2OR —COOH respectively) at the condition of the nucleation medium being studied, require the co-precipitation of counter-ions in the growing crystal to maintain near-zero, or zero, electrical neutrality in the crystal.[98] The electrical considerations of protein crystal growth have also begun to be explored in both experiment and theory.[99-102] Electrostatic forces in crystal alignment have been characterized for liquid crystals.[103] and micro-ion disposition.[104] These findings could be of relevance to protein crystal growth.
A well-known phenomenon that deserves mention here is that the interaction of intact cells with substrates has been studied in detail. Most studies have concluded that electrostatic interactions, at least for cell-substrate interactions, are important.[105-110] The sorption of organic compounds onto inorganic compounds has also be studied extensively.[111, 112] and the study of the range of chemical reactions catalyzed on such surfaces remains an active area of research.[113-123] A related discipline is biomineralization, which involves the use of organic compounds such as proteins to promote or catalyze the formation of solids of inorganic compounds.
Recent observations in the inventors' laboratory suggest that the net excess charge in a nucleation medium (i.e. a reaction vessel such as a levitated droplet) is an experimentally accessible variable that does in fact affect the magnitude of the barrier for nucleation in the condensed phase. These finding are described in detail below.
In accordance with the invention, a method of controllably inducing nucleation of a first solute dissolved in a solution is described. The method includes the steps of providing a primary vessel for containing said solution; applying an induction potential to said primary vessel such that said solution acquires a net charge; and causing ion-induced nucleation of at least some of said first solute in a condensed phase.
The step of causing ion-induced nucleation may comprise maintaining the surface charge density of said primary vessel above a threshold value and/or maintaining the mass-to-charge ratio of said primary vessel below a threshold value. Ions in the vessel in excess of any counterions induce heterogeneous nucleation of the solute.
In one embodiment of the invention the primary vessel may be wall-less. For example, the primary vessel may be a droplet. The surface charge density may be maintained above the threshold amount in an outer portion of the droplet at an air/droplet interface. After an induction potential has been applied to the droplet, it may be levitated. Many different means of levitation may be employed, such as an electrodynamic balance. The solution may comprise a surface tension modifier to inhibit Coulomb explosion of the droplet.
In other embodiments the vessel may be a droplet in combination with a surface. The primary vessel may comprise a portion of a conduit holding the solution. For example, the conduit may be a capillary.
The ion-induced nucleation may cause formation of one or more nuclei. Volatile solvents in the solution are allowed to evaporate to yield a residue comprising the one or more nuclei. Evaporation of the volatile solvent(s) may have the effect of increasing the concentration of the first solute in the vessel. At least some of said nuclei may be used to promote crystallization of the first solute. The method may further comprise the step of delivering the nuclei to a target location. The target location may be a substrate adapted to receive the nuclei. For example, a portion of the solution comprising the nuclei may be deposited on the substrate. At least some of the nuclei may be delivered from the primary vessel to a secondary vessel for seeding crystal growth in the secondary vessel.
The first solute is preferably a solid dissolved in the solution. For example, the first solute may be an inorganic compound or an organic compound. In one embodiment the first solute may be a biomolecule, such as a protein. Inorganic compounds could include metals, melts and alloys.
In one embodiment of the invention a second solute may be dissolved in the solution in addition to the first solute. The method may comprise selectively precipitating the first and second solutes in order to separate and/or purify the solutes. The method steps may be automated for this and similar purposes. For example, the first and second solutes may be stereoisomers or enantiomers. The invention may also be employed to separate one polymorphic form of a compound from another. In one embodiment the second solute may be a MALDI matrix. The method may result in co-crystallization of the first and second solutes. The method also encompasses precipitates and co-precipitates produced by the method steps.
In one particular embodiment of the invention, the method comprises controllably inducing precipitation of selected solutes dissolved in a solution comprising providing a primary vessel for containing said solution; applying an induction potential to said primary vessel such that said solution acquires a net charge; and selectively causing ion-induced precipitation of at least one of said solutes in a condensed phase. The invention also encompasses a method of controllably inducing crystallization of at least one solute dissolved in a solution, said method comprising providing a primary vessel comprising said solution; controllably imparting a net charge on said solution in a condensed phase to selectively cause ion-induced nucleation of said at least one solute; and depositing crystals derived from said nucleation on a substrate.
In drawings which describe embodiments of the invention but which should not be construed as restricting the spirit or scope of the invention in any way,
Throughout the following description specific details are set forth in order to provide a more thorough understanding of the invention. However, the invention may be practiced without these particulars. In other instances, well known elements have not been shown or described in detail to avoid unnecessarily obscuring the present invention. Accordingly, the specification and drawings are to be regarded in an illustrative, rather than a restrictive, sense.
The inventors have previously described various methods for levitation and manipulation of charged droplets. Apparatus used by the inventors for droplet levitation is described in detail in the literature and in Applicant's prior international application No. PCT/CA01/01496 filed 23 Oct. 2001 and entitled “Method and Apparatus for Producing a Discrete Particle” (WO 02/035553 A3), the disclosure of which in hereby incorporated by reference.[8, 124]. Related subject matter is also described in Applicant's prior international application No. PCT/CA04/000242 filed 24 Feb. 2004 and entitled “Formation of Closely Packed Microspots and Irradiation of Same” (WO 2004/075208 A3), the disclosure of which is hereby incorporated by reference. As shown in
Particulars of the droplet generation, levitation and deposition processes are described in detail in the experimental section below. This approach is sometimes referred to as “wall-less sample preparation” (WaSP). Briefly, a small amount of a starting solution of known composition is initially loaded in droplet generator 10. The starting solution may contain one or more solutes of interest (typically present as dissolved solids). The starting solution also typically includes one or more volatile and non-volatile solvents. A quantity of solution is ejected from the nozzle of droplet generator 10 in the vicinity of induction electrode 16 to form an initial droplet. As the initial droplet emerges from droplet generator 10, a DC potential applied to induction electrode 16 induces a net charge on the droplet. The initial droplet may then by deposited on to a substrate directly or injected into electrodynamic balance 12 and levitated there for a period of time.
Typically volatile solvents present in the initial droplet evaporates quickly to yield a residue of the initial droplet. The residue is comprised of solvents and solutes of lower volatility. Coloumb explosion of the initial droplet or droplet residue, a process that ordinarily causes a droplet with net charge to fragment, can be avoided by including in the starting solution a compound having high surface tension, such as glycerol.
Droplet residues having net charge may be suspended in the EDB 12 for a desired length of time (which can vary from a few milliseconds to several hours), for example to allow initiation or completion of a desired chemical reaction, and may then be controllably delivered to a target location, such as substrate remote from the EDB. As explained above, the substrate on which the droplet residue is deposited may be a MALDI plate 18 in one embodiment of the invention. As described in Applicant's prior PCT application (WO 2004/075208 A3), the Applicant's method enables droplet residues, or portions thereof, to be deposited on the target substrate as microspots in a spacially precise manner. For example, different microspots may be deposited on the substrate in very close proximity to one another.
The deposited material may then be further analyzed or characterized by various different means, as described further below. For example, the microspots may be irradiated and the resulting ions detected by mass spectrometry, such as time of flight mass spectrometry. The deposited material may also be characterized using an optical microscope or the like.
The present invention had its genesis in the unexpected observation that solutes (i.e. dissolved solids) present in the starting solution had a greater propensity to nucleate and form crystals when the reaction vessel (e.g. a levitated droplet) was subjected to an induction potential of larger magnitude such that the reaction vessel had a net excess charge. In particular, it was initially determined that the energy barrier for dissolved solids to nucleate is affected by the mass-to-charge ratio of the reaction vessel. A reduction in the mass-to-charge ratio of the vessel causes the barrier for nucleation to decrease. More recently, the inventors have determined that the most relevant experimental factor appears to be the surface charge density of the vessel. That is, there appears to be a linear relationship between the surface charge density of the vessel (e.g. a droplet) and the onset of nucleation. As described in detail below, in droplets having a high surface charge density (i.e. a low mass-to-charge ratio), a large number of small nuclei are observed. In contrast, in droplets having a lower surface charge density (and a relatively higher mass-to-charge ratio), very few or no nuclei were observed in residues of the droplets. When solids did form in those levitated droplets, they tended to be aggregates rather than nuclei. These observations suggest that reaction vessels that have an appropriate surface charge density (and mass-to-charge ratio) promote the formation of nuclei, and possibly catalyze the nucleation.
The sample pictures shown in
The inventors have previously measured the distribution of an organic dye cation (Rhodamine 6G) within NaCl that results when a droplet with net charge is allowed to dry while levitated.[125] From that work, a measure of the thickness of the surface layer that contains the net charge on a droplet with net charge was obtained. That data indicates that droplets with net charge can be described as imperfect conducting spheres. The thickness of the surface layer was determined to be several micrometers in thickness which is much larger than the expected thickness of an electric double layer. This suggests that the surface volume is quite different in its chemical and physical description than the interior of the droplet with net charge. As explained above, the ion induced nucleation phenomenon which varies with the surface charge density (and the mass-to-charge ratio of the reaction vessel) may be as a result of the electric field at the interface between the surface and bulk-like interior volumes within these droplets, plus the fact that electrical neutrality is not maintained in these droplets with net charge. The presence of an electric field (i.e. increased net charge, so reduced mass-to-charge ratio) appears to influence the magnitude of the thermodynamic barrier leading to nucleation of a solute. The inventors believe that the electric field causes the alignment of molecular dipoles of the solutes in the droplet, and that effects a reduction in their internal energy. As explained below, the inventors have demonstrated lowered solubility of some solutes as a function of the net charge of the reaction vessel.
As will be appreciated by a person skilled in the art, many reactions occur as a result of a charge imbalance within or between reaction vessels that have zero net charge. By contrast, as explained above, the present invention relates to reaction vessels having “net charge” or “net excess charge” which are suitable for promoting ion-induced nucleation. The term “ion” as used herein refers to atoms or molecules that carry charge. The terms “net charge” and “net excess charge” as used herein refers to the presence of ions in a vessel of a single polarity that are in excess of the counterions of opposite polarity present within the same vessel. As used in this patent application the reaction “vessel” may simply consist of a droplet of a solution having a net charge, or may alternatively consist of a droplet together with a supporting surface. For example, a “vessel” may consist of a droplet deposited on a surface or held within a container, such as a capillary or portion thereof. Most references herein to the terms “reaction vessel” and “nucleation vessel” refer to a droplet wherein nucleation is induced, but the invention is not restricted to that embodiment.
The inventors' findings can be exploited as described herein in order to elicit selective control over the induction of nucleation and subsequent crystallization of target solutes of interest in the condensed phase. The inventors anticipate that this ion induced nucleation phenomenon, in reaction vessels having a desirable surface charge density, is likely to be general for all dissolved solids, ranging from inorganic compounds, to low and high molecular weight organic compounds, including proteins and other molecules. For example the present invention can be used to selectively crystallize a target solute or to separate different solutes from one another based on their propensity to nucleate at different reaction conditions. The different solutes could constitute different compounds or different stereochemical forms of same compound. The invention could also be exploited to controllably select or separate polymorphic forms of a compound (which may often have very different biological activity). The crystals derived from the process could be the subject of further analysis, characterization or manipulation, for example as a prepared sample material for MALDI-TOF MS. Examples describing the controlled induction of nucleation and crystallization of various compounds are described in detail below.
The following examples will further illustrate the invention in greater detail although it will be appreciated that the invention is not limited to the specific examples.
The following description of experimental details and experimental results is presented in multiple parts. Example 1.0 describes an observation regarding the nucleation of an organic compound (CHCA) in a levitated droplet. The reproduction of that observation led to the realization that the net charge of the reaction vessel (e.g the mass-to-charge ratio of the vessel) influences the nucleation of a solute. In this example the reaction vessel is a levitated droplet. Example 2.0, which summarizes experiments performed independently of Example 1.0, describes measurement of droplet mass and net charge, and the filtering of droplets with net charge (i.e. the reaction vessels in the work described in section Example 1.0) as a function of their mass-to-charge ratio. Example 2.0 also describes the effect of allowing a droplet dispensed from a micropipette to be deposited on to a biased plate. Example 3.0 describes MALDI matrix and analyte compound co-precipitates. Example 4.0 describes promotion of CHCA and peptide cocrystallization within levitated droplets having net charge. Example 5.0 describes the measurement of chemical parameters, such as promotion of NaCl precipitation, in droplets with net charge which were not allowed to undergo Colulomb explosion. Example 6.0 describes ion-induced precipitation of NaCl, CHCA, THAP and samples of D and L serine in levitated droplets possessing net charge.
Apparatus used by the inventors for droplet levitation is described in detail in the literature and in Applicant's prior international application No. PCT/CA01/01496 filed 23 Oct. 2001 and entitled “Method and Apparatus for Producing a Discrete Particle” (WO 02/035553 A3), the disclosure of which in hereby incorporated by reference.[8, 124] As described generally above and as shown in
Droplet generator 10 used in the following examples is a commercially available ink-jet style, droplet-on-demand generator (Microfab, Plano, Tex., USA, e.g models MJ-AB-01-60 and MJ-AB-01-40) which requires as little as 2 μL of starting solution to function. As explained below, the starting solution may include both volatile and non-volatile solvents and solutes, including the solutes targeted for analysis. Each droplet is generated by applying a time-dependent waveform to an annular shaped piezoelectric crystal bonded to the outside of the glass capillary of the droplet generator 10. The size of the piezoelectric crystal change and the time dependence of that crystal size change effected by the amplitude and temporal characteristics of the AC waveform respectively, create a pressure wave inside the glass capillary of the droplet generator 10. In turn, that pressure wave forces a volume of liquid out of the nozzle of the droplet generator. While that volume of liquid is emerging from the nozzle, it takes on the form of a jet, and the DC potential applied to the induction electrode 16 induces a net charge onto that jet of liquid such that when the momentum imparted into the jet causes that jet to separate from the nozzle and the jet collapses into a droplet, that droplet has a net charge. The droplet generator is positioned such that each droplet flies into the center of an electrodynamic balance (EDB), where it can be trapped and levitated, provided the electric field and the droplet's mass-to-charge are appropriate.
As explained above, the starting solution loaded into the droplet generator, is typically prepared by mixing several volumes of different stock solutions together. Stock solutions are used because one or more of the target solutes may require dissolution in a particular solvent. By way of non-limiting example, a starting solution with a total volume of 400 μl was prepared by the addition of: i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume; ii) 40 μL of acetone; iii) 40 μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water; iv) 180 μL of acetonitrile; and v) 80 μL of distilled deionized water. In another example, a starting solution with a total volume of 400 μl was prepared by the addition of: i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume; ii) 40 μL of acetone; iii) 40 μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water; iv) 180 μL of acetonitrile; and v) 80 μL of distilled deionized water. In yet another example, a starting solution that contained a different volume of the saturated solution of CHCA was prepared from: i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume; ii) 40 μL of acetone; iii) [40+(x)] μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water; iv) [180−(0.5×)] μL of acetonitrile; and v) [80−(0.5×)] μL of distilled deionized water. As will be apparent to a person skilled in the art, the identity and concentration of solutes and solvents in the starting solution may vary depending upon the desired analysis.
Induction electrode 16 applies a net charge to a droplet of the starting solution as it emerges from droplet generator 10. By way of example, induction electrode 16 may be made of copper shaped in a disk with a 4 mm diameter hole cut in its center. As shown in
In use, immediately upon formation of the droplet, solvents in that droplet begin to evaporate. For example, one or more of the solvents in the droplet are typically of low viscosity and high vapor pressure. These solvents rapidly evaporate, usually within seconds after formation of the droplet. This solvent evaporation is occurring while the droplet flies to the EDB 12 and continues while it is levitated in the EDB. The starting solutions used in these examples typically incorporate glycerol at a few percent by volume. Glycerol is a solvent of high viscosity and low vapor pressure to avoid Coulomb explosion and enable droplet lifetimes on the order of hours (although in these examples the droplets were often levitated for only a few minutes). The physical and chemical description of the levitated droplet after the rapid evaporation of its volatile solvents is a function of the starting solution composition, the conditions used for droplet generation, whether or not Coulomb explosion occurred, and the environmental conditions such as temperature and humidity in the chamber in which levitation is performed. At this stage in the process the levitated droplet is sometimes referred to as a droplet “residue” in the sense that its composition, though known, is at this stage quite different than the composition of the starting solution prior to evaporation of solvents.
Following a period of levitation, which can range from a few milliseconds to hours, the droplet residue is typically deposited onto a substrate remote from the EDB by adjustment of the electric field in the EDB 12, such as the MALDI plate 18 (
In a first experiment, the number and type of solids in the residues of levitated droplets that had been produced using a positive induction potential (i.e. the droplets had net negative charge) has been observed to vary with the magnitude of the DC potential applied to the induction electrode. The results of this experiment in which the only variable was the DC potential applied to the induction electrode 16 are presented in Table 1 below. (Because this was one of the first reproducible observations of dissolved solids forming in levitated droplet residues as a function of the levitated droplet's mass-to-charge ratio, the inventors did not segregate precipitated solids on the basis of whether they were aggregates or crystals).
In this particular experiment, a single waveform was used to generate the droplets and thus the nominal mass of each of the droplets was unchanged within experimental error. Thus the change in the induction potential caused a change in the droplet mass-to-charge ratio.
The composition of the starting solution (total volume=200 μL) was as follows: i) 30 μL of a solution containing 20% glycerol to distilled deionized water by volume, ii) 50 μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water, and iii) 120 μL of distilled deionized water
The data in Table 1 indicate that there was an increase in the number of solids present in the droplet residues as a function of the DC potential applied to the induction electrode 16 at the time of formation of the droplets. A higher DC induction potential resulted in a more CHCA solids in the residues of the levitated droplets.
Another nucleation experiment was performed as shown in
The levitated droplet residues were each deposited after a levitation period of 3-5 minutes in this experiment. The photographs clearly show that the nucleation barrier was affected (i.e. reduced) when the induction potential was raised. The black appearance of globular shaped aggregates are apparent in
The number of crystals, and the size of those crystals, from 18 different droplets at each of the induction potentials 100, 150, and 200 V were further characterised using an optical microscope as shown in
Within the crystal growing community, it is known that the full description of the state of a system should include an electrostatic term, but because crystallisation has almost always performed within a net neutral solution, the electrostatic term is very often simply neglected as an experimental variable. The inventors have recognized that net charge is in fact a controllable variable in the nucleation of a solute in the condensed phase.
A nucleation experiment was conducted used a starting solution that was composed of i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume, ii) 40 μL of acetone, iii) 100 μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile: 0.1% TFA in distilled deionized water, iv) 150 μL of acetonitrile, and v) 50 μL of distilled deionized water.
This experiment had two main purposes. The first purpose was to study the effect of crystallization in droplets that contain a net positive charge instead of negative. The second purpose was to gain a better understanding of the morphological details of the crystal surface that were apparent using an optical microscope. From previous results, it was learned that droplets that had been collected on a glass slide were solid and dark in appearance (i.e. visually resembled coffee beans) which made it relatively difficult to discern their surface characteristics using an optical microscope. Hence, it was decided to collect the droplets on a stainless steel MALDI plate that was chrome plated. The surface of this plate had been machined flat and polished. Crystals within the residues of the droplets deposited from the EDB 12 appeared more shiny and defined under the microscope. However, the small crystals that were observable when the droplets had been deposited onto a glass cover slip, were not observable on the stainless steel plate, possibly because there was not sufficient light available to illuminate the solids sufficiently to allow viewing of the smaller crystals. The crystals observed in the deposited droplet residues, and segregated according to their size, are presented in Table 2 below.
The numbers of precipitates in each glycerol droplet is noted in Table 2. There were 44 replicates performed using an induction potential of-100 V, and 8 replicates performed using an induction potential of −180 V. The crystals sizes observed were categorized as either >3.5 μm in diameter or 1.0-3.5 μm in diameter.
The data presented in Table 2 provides clear evidence that higher relative DC potentials applied to the induction electrode 16 cause the formation of a larger number of crystals relative to when a lower DC potential is used. This experiment was informative to the inventors because this was the first time that the number of crystals of CHCA formed were classified according to their size, as shown in
The objective of this experiment was to grow a large crystal of CHCA. As described above, the inventors had previously shown that the mass-to-charge ratio of the reaction vessel can be used to preferentially form crystals of CHCA, rather than aggregates. In this example, we studied the potential to grow a single large crystal in a reaction vessel that was prepared with a mass-to-charge ratio that was in the range in which CHCA nuclei were previously observed to readily form. The levitation period of the nucleation vessel was extended so that the kinetics of crystal growth was not a limiting factor in this experiment.
The starting solution used in this experiment was comprised of: i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume, ii) 40 μL of acetone, iii) 40 μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water, iv) 180 μL of acetonitrile, and v) 80 μL of distilled deionized water.
Six droplets were created with the induction potential set at +80V. +80 V is a relatively low induction potential, and mass-to-charge ratio of these reaction vessels were relatively high and hence the inventors did not expect to observe CHCA nuclei in this trial. Three of these droplets were levitated for 3 minutes before being deposited, and the remaining three were levitated for a total of 12 minutes before they were deposited. No solids (i.e. no nuclei, aggregates or crystals) were observed in the residues of any of these droplets.
Next, 8 droplets were created at +180 V. +180 V is a relatively high induction potential, so the mass-to-charge ratio of these reaction vessels was relatively low and hence the inventors did expect to observe CHCA nuclei in this trial. Four of these droplets were levitated for 3 minutes, and the remaining 4 droplets were levitated for a total of 12 minutes before they were deposited. In one of the droplet residues that had been levitated for 12 minutes, there was a crystal that had a length of 21 μm (
A further experiment was conducted to test the hypothesis that nuclei versus aggregation of CHCA could be unambiguously controlled by setting the mass-to-charge ratio of the reaction vessel (droplet). Thus, the hypothesis was that the barrier for nucleation of crystals versus aggregates can be reduced in reaction vessels that have a low mass-to-charge ratio.
The starting solution was prepared by the addition of: i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume, ii) 40 μL of acetone, iii) 40 μL of a solution saturated in α-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water, iv) 180 μL of acetonitrile, v) 80 μL of distilled deionized water.
The solids observed in the levitated droplet residues were classified as aggregates or nuclei. The nuclei were further differentiated by size and three size ranges were used, as indicated in Table 3 below. Large (>3.5 μm in diameter), medium (1.0-3.5 μm in diameter), and small (<1.0 μm in diameter)
Table 3 shows nuclei (either 1.0-3.5 μm in diameter or <1.0 μm in diameter) versus aggregates (agg) in levitated droplets as a function of the induction potential, either 100 or 190 V, used to induce net charge during the formation of each droplet. The number of replicates performed with an induction potential of 100 V and 190 V was 12 and 6 respectively.
When solids were formed in these levitated droplets, only aggregates were observed in the droplets that were formed using an induction potential of 100 V (i.e. a relatively high mass-to-charge ratio for the reaction vessel), but in contrast, only nuclei were observed in the droplets that were formed using an induction potential of 190 V (i.e. a relatively low mass-to-charge ratio for the reaction vessel).
A further experiment was performed to test the hypothesis that nuclei formed in a primary reaction vessel at low relative mass-to-charge could be delivered to a secondary nucleation vessel containing another solution to seed the formation of crystals.
The solution used for droplet generation and subsequent levitation of droplets with net charge was comprised of: i) 60 μL of a solution containing 20% glycerol to distilled deionized water by volume, ii) 40 μL of acetone, iii) 40 μL of a solution saturated in a-cyano-4-hydroxycinnamic acid (CHCA) in 50:50 v:v acetonitrile:0.1% TFA in distilled deionized water, iv) 180 μL of acetonitrile, and v) 80 μL of distilled deionized water. The solution pipetted directly onto the glass slide consisted of a saturated matrix solution of CHCA in 1/1:v/v: ACN/0.1% TFA in distilled deionised H2O). The volume of this solution pipetted in each case was 50 μL.
To provide a visual reference for CHCA, the solids formed from a solution simply deposited onto a glass cover slip are shown in
This experiment can be classified as a two-step crystal design experiment, wherein the “primary reaction vessel” could be optimized for the formation of small nuclei of a dissolved solid (i.e. tuning of the chemical and physical description of a levitated droplet), and the “secondary nucleation vessel” that is seeded with nuclei could be optimized for crystal growth.
The capability to process compounds in levitated droplets with net charge, such as promoting or catalyzing the nucleation of dissolved solids contained within them was demonstrated in Example 1.0. In this Example the inventors describe how droplets with net charge levitated in an EDB 12 can be filtered according to the mass-to-charge ratio of the droplets. This section also delineates measurements of the actual mass and net charge in individual droplets.
As described above, and shown in
All charged droplets studied in this Example were prepared using a piezoelectric droplet-on-demand dispenser 10 as described in Section 1.1.1 above fitted with a 60 μm diameter orifice. Methanol or aqueous solutions containing 0.8 to 8.0% glycerol were loaded into the ˜2 μL internal reservoir of the droplet dispenser using a 10 μL automatic pipette. Droplet formation and charging characteristics are described in detail below. The charge carried by single droplets was measured by delivering the droplets to a stainless steel plate connected to an electrometer (6517a, Keithley Instruments). A Faraday cage was required to reduce the background level so that the small charges (femtoCoulomb) carried by each of the droplets could be measured.
The EDB 12 used for wall-less sample preparation (WaSP) has been described in detail earlier in Section 1.1.1. In this Example the EDB 12 consisted of two copper wire (0.9 mm diameter) rings 14 (2 cm diameter) mounted parallel at a separation distance of 6 mm.[8] The amplitude of the 60 Hz AC potential applied to the ring electrodes (ACtrap), in phase, ranged from 500 to 2,700 V0-P. The vertical positions of droplets in the EDB 12 were manipulated by the DC potentials applied to the induction electrode 16 and the MALDI plate 18. Droplets levitated in the EDB 12 were illuminated via forward scattering by a 4 mW green HeNe laser (Uniphase model 1676, Manteca, Calif.). Images of levitated droplets were collected by focusing a digital camera through a microscope objective to the center of the ring electrodes. Note that to minimize the disturbance of the trajectories of levitated droplets, the droplet generator 10, the ring electrodes 14 of the EDB 12, and the target plate 18 were enclosed in a plexiglass chamber to eliminate excessive air flow (
Single droplets were dispensed directly into liquid scintillation vials using time dependent waveforms applied to the piezoceramic droplet dispenser 10. 2 mL of scintillation cocktail (Amersham Biosciences) was added to the scintillation vial that was then vortexed for 30 seconds. Scintillation is the name given to the detection of fluorescence emission from a compound that was itself excited into an electronic state above the ground state by the absorption of a thermalized electron and the thermalized electron was the result of quenching a hot electron generated by a nuclear decay event. The detection of the fluorescence emission originating in the scintillation vial enables quantitative measurement of the total number of radionuclides in the scintillation vial.
All activity measurements from single droplets and stock solutions were performed using a liquid scintillation counter (LKB Wallac 1217 RackBeta, Fisher Scientific, Montreal, Quebec). Each measurement had an integration time of 10 minutes. Working with 32P requires many safety precautions including the use of plexiglass shielding, a Geiger-Mueller counter, and swipe tests in the working area to monitor for possible spills.
In order to demonstrate the charged droplet filtering capabilities of WaSP, the inventors first established a reproducible method for creating standard charged droplets of known m/z. The basic method is set forth in Example 1 above.
As described above, droplets composed of a low volatility solute such as glycerol have long been the vehicle of choice for EDB applications. To create the glycerol droplets, a solution of 3% glycerol in methanol was loaded into the droplet dispenser's internal reservoir (
To measure the net excess charge induced into the single dispensed droplets by the induction electrode 16, each droplet was captured on a stainless steel plate. Upon impact of the droplet, the charge delivered to the plate was measured using an electrometer with femtocoulomb (fC) sensitivity. The charge on single droplets prepared with ±10 V0-P (the minimum required to create a droplet) applied to the piezoceramic in the droplet dispenser delivered too little charge to be measured one at a time so 100 droplets were dispensed at 100 Hz. The total charge delivered was measured and divided by 100 to calculate the average charge carried by single droplets. This experiment was repeated 20 times (2000 droplets) for four different IPf.
To investigate the reproducibility of single droplet charging, the amplitude applied to the piezoceramic in the droplet dispenser was increased (30 V0-P) to create larger droplets. This enabled the measurement of the charge carried by single droplets because the droplets were bigger (see mass measurements below) and thus carried more charge. At a fixed IPf, twenty droplets dispensed at 0.5 Hz and the total charge delivered to the plate was measured after each single droplet impacted with the plate. By repeating this experiment for IPf set at 9 different values (25, 50, 75, 100, 125, 150, 175, 200, 225 V), droplet charging was deemed reproducible by the linearity of the plot of the total charge delivered as a function of the number of droplets dispensed (
To calculate the m/z of the droplets created, it is necessary to know the mass of the droplets. A radiochemical method was used to measure the volume of droplets as a function of amplitude applied to the piezoceramic in the droplet dispenser, allowing us to calculate the mass of the droplets. A 100 μL solution of 3.7 MBq 32P labeled orthophosphate in a mixture containing 89.2% water, 10% methanol, and 0.8% glycerol was prepared. By depositing single droplets into liquid scintillation vials and comparing the number of Becquerels measured from them relative to the number measured from a 1.000±0.005 μL aliquot of the 32P in glycerol stock solution, we could determine the volume of the single droplet delivered.
The average activity measured from the single droplets was 0.026±0.001% that measured from the 1.000 μL sample so the average initial droplet volume was 2.6×104±1×10−5 μL or 260±10 μL. The solution loaded into the droplet dispenser to create the droplets contained 0.8% glycerol by volume so the initial droplets dispensed contained 2 μL of glycerol. When the droplets are levitated in the EDB 12, volatile solvents rapidly evaporate to leave behind a droplet residue composed of essentially 100% glycerol. (Corrections to account for the water in the droplet that does vary with the relative humidity of the air in the levitation chamber were made subsequently to these calculations). Therefore, the final droplet volume was ˜2 μL. By assuming that the droplet was 100% glycerol and using the density of glycerol (1.259 g/ml), the mass of the levitated droplets on average was calculated to be 2.5×10−9 g. Converting this to atomic mass units (u), the droplets were 1.5×1015 u. Recalling that the glycerol droplets carried charges ranging from 5 to 1000 fC, the m/z of the levitated droplets ranged from 9.6×108 to 2.0×1011 u/e.
Droplets of many different compositions can be prepared using WaSP so it is useful to look at the effect of composition on their m/z. If the percent glycerol in the starting solution was changed to 8%, the final levitated droplet residue would have a volume of ˜21 μL, thereby increasing the mass (and m/z) by 20 times. The inventors have successfully levitated droplets created from solutions containing a percent glycerol by volume of 0.8 to 10%, creating final levitated droplets of 1×1015 to 20×1015 u. The inventors have also shown that the charge carried by the droplets can also be varied from 104 to 106 e. By combining the control of the mass and charge of the glycerol droplets, a wide variety of possible experimental scenarios has been established. Note that these values are many orders of magnitude greater than the typical m/z of ions measured in a quadrupole ion trap mass spectrometer. This characteristic will prove valuable in creating unique applications for this technology in the future.
Using the standardized charged glycerol droplets described above, experiments designed to elucidate whether WaSP could be used to filter the droplets based on their m/z were performed. The general experimental approach was to trap two droplets simultaneously, one created with high m/z and one low m/z relative to each other. By measuring the DC potential applied to the target plate used to eject each droplet, the filtering capabilities of WaSP could be evaluated. An important factor in this experimentation was differentiating between the high and low m/z droplets once trapped. To do this, both high m/z (IPf=50 V) and low m/z (IPf=100 V) droplets were levitated alone (ACtrap=1600 V0-P) and their respective trajectories were compared. Once trapped, the low m/z droplet immediately adopted a trajectory elongated along the z-axis whereas the high m/z droplet was more focused towards the null point (the center of the EDB). When levitated simultaneously, the high and low m/z droplets adopted trajectories similar to when they were levitated alone so this property was used to identify them (
To levitate both high and low m/z droplets simultaneously, ACtrap was set to 1600 V0-P, a high m/z droplet was levitated, and then a low m/z droplet was injected in to the EDB. Attempts to reverse the order of injection resulted in failure to capture the high m/z droplet. When ACtrap was decreased to 1000 V0-P, the trajectory of the low m/z droplet became focused towards the null point, whereas the trajectory of the high m/z droplet moved closer to the induction electrode (
Table 4 below summarizes the DP measured for each droplet in a series of droplet pairs whose relative m/z varied. The DC potential applied to the MALDI plate required to eject each droplet of a pair of droplets created with varied induction electrode potentials is shown.
Each value in Table 4 is the average of the values measured for five separate pairs of droplets and the errors indicate the magnitude of one standard deviation. Pairs of droplets created at identical IPf (identical m/z) were ejected from the EDB at the same DP, within experimental error. However, the second droplet deposited always required a slightly higher average DP. With droplets created at different IPf (different m/z), the droplet with the highest m/z was always deposited first. A higher ACtrap was required to trap the lower m/z droplets so the DP of a droplet that was created with IPf=50 V and trapped with the droplet created with IPf=100 V was lower than when it was trapped with a droplet created with IPf=150 V. This was a direct example of the effect of ACtrap on DP we have observed in the past. Overall, the preferential ejection of the highest m/z droplet upon increase of the target plate potential indicated that m/z was a factor in ejection order and hence the EDB essentially acted as a mass filter for the charged droplets.
As should be apparent from this Example, there are two ways that WaSP could be used to filter droplets. The first would be to trap a population of droplets, anywhere from 1-50 droplets, in the EDB. Then, single droplets can be ejected from the balance onto a target based on their m/z as described above, while the remaining droplets remain trapped. If desired, this stably levitated population could then be exposed to gas-phase reactants to modify their chemistry before their ejection, enabling experimental determination of their environment on the ability to identify them. The second approach would be to set up the EDB to act as a bandpass filter, allowing only a certain range of m/z to pass through it, automatically ejecting the droplets in less than 500 ms. This method would be useful for rapidly sorting of droplets based on their m/z and could be readily automated.
During the development of these modes of operation the inventors used a translatable collection plate to make the demonstration clearer. In the first mode, the delayed ejection mode (
There are several aspects of WaSP charged droplet filtering that make it very flexible with respect to the types of application for which it could be used. First of all, the droplets that are filtered by WaSP are created from a starting solution of choice, therefore putting very little limitation on the potential analytes studied, such as single bacteria or synthetic inorganic particles. Next, the m/z of the droplets does not have to fit within the range that was demonstrated in this Example. By changing the characteristics of the electric field, WaSP can be easily modified to filter alternate m/z's. The source of the charged particles is also not limited to strictly charged droplet dispensers. The WaSP approach is potentially applicable to filter atmospheric particles using alternate charging mechanisms. Lastly, the target for particle collection is not limited to a stainless steel plate. The inventors have also delivered particles onto populations of cells on a glass slide and into the orifice of a mass spectrometer.
Overall, the flexibility of this charged particle filtering methodology based on WaSP has proven valuable in the development of such applications as the preparation of μm-sized sample spots for MALDI-TOF-MS and the study of multi-phase heterogeneous reactions on single aerosol particles. Some other potential applications of an using an EDB as a particle filter include: (1) a delivery module for a bioaerosol mass spectrometer attempting to detect single cells, bacteria, or viruses, (2) isolation of inorganic particles for subsequent studies based on their m/z, (3) an aerosol particle sorting mechanism, (4) a tool for performing studies of aerosol particle reactivity or nucleation capabilities based on their m/z, or (5) the single particle dosing of cell populations for medical research.
The rationale for performing the experiments described in this section was to confirm that the potential applied to the plate being used as the target for deposition of a levitated droplet/particle was not influencing test results.
In the above Examples, the nucleation of CHCA was described as being influenced by the mass-to-charge of the reaction vessel (e.g. in a droplet). However, in all of those experiments, the extent of nucleation was not measured in situ (i.e. in the levitated droplet) but rather only after the levitated droplet had been deposited on a target. The target was either a stainless steel plate to which a potential was applied to attract the levitated droplet or a glass cover slip positioned on top of the stainless steel plate to which a potential was applied. To eliminate the possibility that the deposition of the droplet from the EDB onto a substrate was responsible for the observations made above, the inventors proceeded to pipette an aliquot of a stock solution onto a biased stainless steel plate. Note that in these experiments, even though the plate is charged, the droplet deposited onto it remained net neutral because within the droplet a double layer would be established immediately upon contact with the plate. So, even though the actual material deposited onto a plate in this section was a neutral droplet, and thus different from the deposition of a droplet with net charge, the effect of the deposition onto a biased plate on the nucleation of dissolved solids was the objective of this experiment. In other words, the hypothesis was, does the drying of a droplet on a biased plate affect crystal formation and growth, and does that in turn affect signal-to-noise ratio (S/N) of the analytes contained in the droplet?
A stock solution of 10 pmol/μl of renin was prepared in distilled deionized water with 0.1% trifluoroacetic acid. The MALDI matrix α-cyano-4-hydroxycinnamic acid (CHCA) was prepared at 10 mg/ml in 50:50 methanol:acetic acid. 10 μl of each solution was mixed in a microcentrifuge tube and vortexed. The MALDI plate was connected to a DC power supply of +500 V. Four 1 μl aliquots were deposited on to the MALDI plate as four discrete sample spots and allowed to air dry for 15 minutes. The MALDI plate was then grounded and four more 1 μl aliquots were deposited onto the MALDI plate and allowed to air dry for 15 minutes. Images of the sample spots were collected through an optical microscope at 4 and 10 times magnification (
The plate onto which the sample was aliquotted using a micropipette was biased to 1000 V, and the experiment repeated. The analysis of the data acquired using MALDI-TOF-MS of the sample materials prepared on a grounded plate and a plate biased to 1000 V during the period that the droplets dried are presented in
Examination of the data presented in
This Example illustrates of how the invention could be applied to the growth of crystals of a MALDI matrix within which an analyte compound co-precipitates. Prior work from other groups has indicated that homogeneous co-crystallization of a solute with a matrix compound is an important factor for obtaining good signal-to-noise (S/N) ratio for the analyte when characterized using MALDI-TOF-MS.[126] Furthermore, the quality of the crystals of a matrix that form on a surface has been characterized as heterogeneous.[127-128] This body of information suggests that there is a need for methodology that enables the formation of good quality crystals within which there is an analyte distributed homogeneously within that crystal in order to realize adequate analytical performance, such as analyte S/N, when characterized by MALDI-TOF-MS.
Several sets of experiments were performed in which droplets were created using different induction potentials. Each experiment was performed with a starting solution prepared daily with the following composition: 40 μL Acetone, 100 μL of a CHCA solution that was originally prepared by mixing 1:1 v:v acetonitrile:0.1% TFA in distilled deionized H2O, 60 μL of water/glycerol solution that was originally prepared by mixing 1:4 glycerol:distilled deionized H2O, 50 μL of ACTH solution with a concentration of 10 μM which makes the final concentration in the solution used for levitation 1 μM, and 150 μL Acetonitrile (ACTH=Adrenocorticotropic Hormone, Fragment 18-39). All droplets were levitated for a period of 5 minutes prior to their deposition onto a substrate remote from the EDB. The substrate was a stainless steel MALDI plate in this example.
The MALDI plate was then inserted into MALDI-TOF-MS (model Voyager, Perceptive Biosystems, MA) and the levitated droplet residues targeted by the laser and the mass spectra obtained are displayed in
Based on the data presented in these figures, the utility of a high induction potential during the droplet formation results in an improved S/N ratio for an analyte compound in the starting solution.
Procedures for dispensing droplets with net charge, measurement of initial droplet net charge and volume, droplet levitation, and matrix-assisted desorption/ionization mass spectrometry are described above. In this experiment promotion of MALDI matrix/peptide cocrystallization in levitated droplet residues was demonstrated.
Two starting solutions were used in the cocrystallization studies. A single component peptide solution was prepared by mixing 40 μL of acetone, 100 μL of a saturated solution of CHCA in 50:50 acetonitrile (ACN):0.1% trifluoroacetic acid (TFA) in H2O, 60 μL 20% glycerol in H2O, and 50 μL of 10 μM adrenocorticotropic hormone fragment 18-39 (ACTH) in H2O (0.1% TFA). A multi-component peptide solution composed of 40 fmol/μL angiotension II and bradykinin, 4 fmol/μL angiotension 1, 0.2 mg/ml CHCA, 20% methanol, 20% ACN, 1.5% glycerol, and 0.6% TFA in H2O was also prepared. An aliquot of the starting solutions was loaded into the approximate 5 μL reservoir of the droplet dispenser using a 10.00 μL automatic pipette.
The single component peptide starting solution was used to generate 10 droplets that were each levitated for 2 min prior to depositing all of them at a single location on a glass slide to create a single sample spot. Optical microscopy of the sample spot composed of 10 droplet residues showed that CHCA precipitates were created during levitation. Insets a-d of
The data presented in
Inducing varied net charge on dispensed droplets of constant volume changed their respective m/z. To effectively study droplets of differing m/z using the EDB, it was important to delineate whether droplets of different m/z could be differentiated from each other while levitated in the EDB. First, an individual IPf,150V droplet was levitated. After noting its trajectory, it was ejected and then a single IPf,150V droplet was levitated at the same ACtrap. Different trajectories for these two droplets were discernible to the naked eye. Next, an IPf,50V and an IPf,150V droplet were levitated simultaneously in the EDB. Again, their trajectories were clearly different. This observation enabled repeat experiments to be performed on two different m/z droplets simultaneously levitated in the EDB because it was possible to visually track which droplet was which.
Another way to differentiate between the two droplets of different m/z was developed when the DC potential applied to the target required to eject each droplet was measured. For all repetitions of the experiment, the IPf,50V droplet was ejected from the EDB first and the IPf,150 V droplet remained levitated. This observation was consistent with all combinations of two different m/z droplets levitated in the EDB (i.e., IPf,150V droplet ejected before IPf,150V). With only two droplets that had the same m/z levitated, such as two IPf,50V, IPf,100V, or IPf,150V droplets, the trajectories of the droplets were similar and their deposition potential (DP) values were identical within experimental error [124]. Thus, two separate and complementary procedures were developed to differentiate between two different m/z droplets levitated in the EDB, facilitating further experiments designed to characterize differential chemical processing as a function of droplet net charge.
Next, the complexity of the experiment was increased. A population of five droplets was injected into the EDB where each had been created at IPf=100, 125, 150, 175, and 200 V, respectively. The DC potential applied to the induction electrode was adjusted to 0 V after 2 min of levitation time. The voltage applied to the target was then manually ramped to higher values and DP was measured for each of the five levitated droplets. This procedure was repeated for nine more sets of five droplets. A plot of mean DP versus IPf was linear over the range IPf=100 to 200 V, indicating that none of these droplets underwent a Coulomb explosion (
The histogram shown on the left hand axis in
Next, the number of droplets in the EDB was increased to 10 by injecting two droplets created at each distinct IPf (
Using an aliquot of the multi-component peptide starting solution to load the droplet generator, 20IPf,50V droplets and 20 IPf,200V droplets were created (and injected) sequentially into the EDB and those two populations levitated simultaneously (
By inserting the MALDI plate into a vacuum chamber (˜3×10−7 torr), the glycerol from these spots evaporated, facilitating observation of the solids that had formed in the levitated droplets by light microscopy (inset 3 and 4,
Another replicate of the experiment was performed to prepare fresh sample spots. A sample spot was created from 12 codeposited IPf,90V droplets and another sample spot was created from 12 codeposited IPf,170V droplets. Each of these spots contained 6 fmol ACTH. The laser output was directed at each of the sample spots and the laser was fired 50 times at the lowest irradiation setting. The ions detected from this set of laser shots were accumulated as five mass spectra (ten laser shots each), and from that, the mean peak height of the monoisotopic ion, [ACTH+H+] at m/z=2465.2 was calculated and plotted with an error of one standard deviation of the mean. The laser irradiation setting was incremented and the analysis repeated (
The importance of net charge in the medium during nucleation versus placing an electrode in a medium to raise its potential above ground was investigated. Aliquots of the single component peptide starting solution (1.00 μL aliquot, 319 fmol/μL of ACTH deposited) were pipetted directly onto a MALDI plate and allowed to dry while the MALDI plate was either grounded or biased with a DC potential. Note that in these experiments each aliquot was overall net neutral, and the aliquot spread over a region circular in shape and had a diameter of 1.5 mm once dry. Within the area irradiated by the N2 laser that was focused to a spot size of 200 μm in diameter, there would have been approximately 6 fmol of material. When the [ACTH+H+] signal intensity was measured from these sample spots that were large relative to the laser spot size, the signal intensity increased and then stabilized at a plateau (
In this Example the following chemical reagents were used: 3.7 MBq 32P labelled orthophosphate, NaCl, and 20 nm diameter fluospheres (Molecular Probes, Invitrogen Inc., Burlington, ON, Canada). These fluospheres are polystyrene based spheres (density =1.05 g ml−1) that encapsulate ˜180 fluorescein molecules per fluosphere. The composition of all starting solutions from which single droplets were dispensed consisted of a single solute, either NaCl or the fluospheres, dissolved in either 100% distilled deionized water or in 97:3 distilled deionized water:glycerol. The 3.7 MBq 32P labelled orthophosphate was used only to determine the volume of starting solution dispensed in discrete droplets.
For droplet dispensing a micropipette was used to load a 3-5 μL aliquot of a starting solution into the reservoir of an ink-jet style droplet dispenser as described above. The nozzle of the droplet dispenser was aligned overtop a 5 mm diameter hole cut in a flat electrode, and positioned with a separation distance of 2 mm between the nozzle and the electrode. A dc potential applied to this electrode established an electric field between it and the nozzle of the droplet dispenser. The electric field influenced ion mobility in the volume of liquid that would become the droplet, but only while that volume remained in contact with the bulk liquid inside the reservoir of the dispenser. The induced charge separation within that volume of liquid caused the resultant droplet to have net excess charge as described above. Each droplet passed through the hole in the induction electrode, and into an EDB where it was then trapped and levitated.
For the purpose of measuring the magnitude of the net charge carried by individual droplets, they were dispensed directly onto a metal target plate connected to an electrometer (model 6517a, Keithley Instruments, Cleveland, Ohio). For these measurements, the droplet dispenser and metal target plate were situated inside a Faraday cage. dc potentials were applied to the induction electrode in the range from 100 to 200 V. For induction potentials of 100, 150, and 200 V dc, the induced net charge per droplet were measured to be −135±11, −235±12, and −325±18 fC, respectively, with a starting solution containing 100 mM NaCl in distilled deionized water loaded into the droplet dispenser.
The initial volume of the droplets dispensed was determined by dispensing a starting solution containing 3.7 MBq 32P labelled orthophosphate directly into a liquid scintillation vial. Radionuclide decay was measured using a liquid scintillation counter (LKB Wallace 1217 RackBeta, Fisher Scientific, Montreal, PQ). Droplets dispensed from the 40 μm diameter orifice had initial volumes of 230±40 μL (average radius 38±2 μm). Droplets dispensed from the 60 μm diameter orifice were measured to have initial volumes of 780±40 μL (average radius 57±2 μm), and this dispenser was used only in the time-lapsed nucleation experiment, the results of which are presented in
The EDB used in this Example has been described above. This EDB was assembled using two ring electrodes and two end-cap electrodes. The ring electrodes were fabricated using 1 mm diameter copper wire that was shaped into 2 cm diameter rings. The rings were aligned parallel with respect to themselves and mounted with a separation distance of 6 mm. This pair of ring electrodes was mounted either parallel or tilted at an angle of 15 degrees relative to the end-cap electrodes.
A 60 Hz sine wave, 0-2,500 V0-P was applied to these rings in phase using a Variac-controlled voltage amplifier that had been constructed in-house. The upper end-cap electrode served two purposes in these experiments. It was the induction electrode during droplet dispensing and it was the top end cap for the EDB during droplet levitation. The bottom end-cap of the EDB also served two roles. A dc potential was applied to it to assist in balancing the droplets at the null position of the EDB, and it also served as the target plate onto which the levitated droplets were deposited at the end of each levitation experiment. The null position of the EDB was defined as a point midway between the two ring electrodes when the EDB was viewed from the side, and when the EDB was viewed from the top, that same point was at the center of the ring electrodes. Adjustment of the dc potential applied to the bottom end-cap created an electric field that imparted a force on the droplet causing it to leave the EDB and impact on the target plate. The bottom end-cap was mounted onto a single-axis translation stage to permit precise relocation of the target plate relative to the ring electrodes of the EDB during an experiment.
Each replicate within an experiment commenced with a droplet dispensing event. The droplet flew into the EDB where it was captured and levitated at the null position of the EDB. The volatile solvent used in the starting solution, distilled deionized water, evaporated rapidly from the droplet, typically within 2 seconds of the droplet formation event [129], leaving behind all solutes and solvents of low volatility. Experiments were performed at relative humidity of 38% at which glycerol to water ratio was 85:15 in those droplets that contained glycerol [130]. As soon as the droplet was dispensed, the dc potential applied to the induction electrode was changed to 0 V in all experiments.
For starting solutions that contained 3% glycerol by volume, the droplets that remained levitated following the rapid loss of volatile solvent were stable with respect to Coulomb explosion. For reference, a net charge of −325 fc contained within a 10 μL glycerol-water droplet is only ˜15% of the net charge that such a droplet could retain before undergoing Coulomb explosion. These droplets could be levitated for periods up to 8 hours, and as such, their ejection from the EDB within this period of time necessitated adjustment of the dc potential applied to the target.
The droplets created from starting solutions that did not contain glycerol underwent Coulomb explosion within 2 seconds of the time of their creation. In these experiments, as soon as each droplet was observed inside the EDB, as verified by laser scatter, the amplitude of the ac potential applied to the ring electrodes was reduced from 2500 to 700 V0-P. The reduced ac amplitude caused the droplet to be levitated near the null position of the EDB with an amplitude of motion less than 0.5 mm prior to it undergoing Coulomb explosion. The explosion itself was characterized by a sudden onset of oscillation of the droplet with amplitudes of ±2 mm from the null position of the EDB. [131] Following the Coulomb explosion, the main residue was observed to levitate briefly (<0.5 s) at the null position and then fall to the target plate.
In most experiments, a glass coverslip was positioned on top of the deposition target, and the levitated droplets were caused to deposit onto it. The deposition of droplets onto a glass target greatly facilitated their subsequent characterization by optical microscopy (Motic, B5 Professional, Richmond, BC) and fluorescence microscopy (Zeiss, Axioplan2, Germany).
Initial experiments involving levitated droplets having net charge were designed to not allow the droplets to undergo Coulomb explosion. Individual droplets were created having either −135 or −325 fC of net charge using a starting solution of 285 mM NaCl in 97:3 water:glycerol. Within seconds of the droplet generation event, the droplet volume had decreased to ˜10 μL. The droplets were levitated for a total of 5 minutes prior to being deposited onto a glass coverslip. NaCl precipitation in levitated droplets was selected for the initial studies because of the relative simplicity of the system, the certainty with which the identity of the ions of the droplet's net excess charge (ionsNEC) were Cl−NEC because of their high abundance relative to other impurity electrolytes and compounds in the starting solution, and because NaCl in biological sample types often suppresses analyte ion signal intensities as measured by mass spectrometry. [132].
Alteration of the magnitude of the net charge imparted onto droplets had a profound effect on the morphology of NaCl precipitates formed during levitation. When droplets contained net charge of −135±11 fC, the resultant precipitate (NaCl(s)) in 23 droplet residues were cubic, except one in which the NaCl(s) had a cube-like morphology (
In 95% of these droplet levitation trials, NaCl(s) was observed to have formed during the period of levitation, and no new NaCl(s) was observed during the post-levitation period when the droplet's residue was in contact with the surface of the glass slide for up to 4 hours after the time of deposition. In the remaining 5% of the trials, no NaCl(s) was observed to have formed in the droplet while it had remained levitated. In these situations, precipitates that did form after deposition were a result of heterogeneous nucleation at the interface between the glass slide and the droplet. The growth of the crystals at the droplet-glass interface resulted in visibly different morphology for those precipitates relative to NaCl(s) that had formed within the levitated droplet, which eliminated ambiguity between precipitates that had formed in the levitated droplets versus in droplets that were in contact with a glass surface.
Comparison of the residues in the images shown in
These trials of NaCl(s) formation in levitated droplets were repeated using the 285 mM NaCl in 97:3 water:glycerol starting solution, except that the droplets were levitated in an atmosphere of N2 by purging the levitation chamber with ultrahigh purity N2. The purpose of the N2 atmosphere was to reduce the possibility that the nucleation being observed in these levitated droplets was due to electrostatically charged dust particles suspended in the atmosphere and that had adsorbed onto the levitated droplets.52 The morphology of the NaCl precipitates that formed in the levitated droplets in this experiment, were similar to that already described and presented in
The results presented in
Possible factors responsible for these observations concerning NaCl nucleation and precipitate growth are the magnitude of the droplet's net charge and the relative rates of solvent evaporation from the droplets. [134] These two factors are related because droplets with higher net charge have higher mobility, and differential mobility of droplets levitated in the EDB would lead to proportionally different rates of solvent evaporation. An indication of the relative role that these two factors have in determining the nucleation and growth of the precipitate was learned from the following experiment. The starting solution consisting of 285 mM NaCl in 97:3 water:glycerol was re-loaded into the droplet dispenser and droplets dispensed using different dc potentials applied to the induction electrode. Each data point in
In addition to the results plotted in
Two prior independent studies investigated nucleation of electrolytes in electrodynamically levitated aqueous droplets. The authors of these reports concluded that small variations of the magnitude of the net charge did not observably affect nucleation. [136,137] The median m/z for the aqueous droplets used in those prior studies are estimated to have been 3.4×109 52 and 7.4×1010. [137] Comparison of the Cohen results to that presented in
Synergistic macroscopic and nanoscopic interactions are likely responsible for the promotion of nucleation (
The promotion of nucleation within droplets having net charge was observed to be dependent on the magnitude of that net charge. The inventors term this phenomenological result ion-induced nucleation in solution. Ion-induced nucleation in the gas phase is the promotion of cluster growth by introducing discrete charged entities that act as nuclei for condensation of vapours, and it has been studied extensively since Wilson reported the phenomenon over one hundred years ago.[143,144] Since then, it has been learned that gas phase ion-induced nucleation occurs because there is a reduction in the free energy required for nucleation and growth relative to that in the absence of an ion.[145-147] For instance, the low pressure synthesis of diamond is rendered favourable over graphite because gaseous C− reduces the critical radius for crystal nucleation.[148] and the net charge of the resulting cluster promotes further growth of the diamond nucleus.[149]
Ion-induced nucleation in solution could have implications for natural phenomena such as nucleation in suspended atmospheric particles, particularly those that contain net charge such as sea salt droplets.[150-151] With further experimental characterization and a theoretical description of this phenomenon, it could find utility as a new tool for laboratory studies of crystal nucleation and growth, or as a medium for the production of nuclei that would be used to seed crystal growth in secondary vessels.19 With these considerations in mind, further development of the methodologies reported herein to enable quantitative characterization of chemical processes that occur in media with net charge are being pursued.
Recall that droplets generated in an ES have an estimated initial m/z of 1×109. We suggest that the m/z of droplets generated by an ES are either at or below the threshold for nucleation at the instant of their formation, and if not, they certainly reach that threshold as suggested by the observation of clusters ES-MS. Furthermore, the promotion of nucleation was observed at magnitudes of net charge less than that required for droplet Coulomb explosion. It is very interesting to point out that in an ES, the pathway that leads to the production of gaseous ions involves the production of droplets having higher relative net charge because of “charge enrichment” in the matter ejected by a droplet Coulomb explosion event.[152,153] Those droplets all existed in a strong electric field, meaning that their increased mobility will also lead to increased rate of solvent evaporation. Hence, this phenomenon of ion-induced nucleation and factors that promote it; magnitude of net charge and possibly also local fluctuations of solute concentrations because of rapid solvent evaporation, all work to promote nucleation and growth of the precipitate. For instance, it could be argued that the recent results from lavarone et al. were in fact a clever example of how to apply ion-induced nucleation in solution to reduce the concentration of the interfering species, while allowing the sought for analyte compounds to remain in the solution phase in the time preceding droplet Coulomb explosion.[154] Interestingly, Julian et al. proposed that magic numbers of serine clusters form by self-assembly in the droplet bulk and later become ionized in the diffuse layer at the droplet-air interface of droplets with net charge.[155] but further studies of this phenomenon by Takats et al. prompted them to speculate that the most abundant serine cluster, the octamer, could itself be formed at droplet-air interface.[156] Furthermore, images of droplet residues reported by Hanton et al. that were obtained by introducing solutions containing organic acids commonly employed as a matrix in MALDI-TOF, such as 2,5-dihydroxybenzoic acid and dithranol, to an ES source for the purpose of preparing films of these compounds.[157] Using scanning electron microscopy to study the resultant films, spherical residues were observed, as expected based on complete evaporation of solvent from the droplets, but, many of those spheres were hollow. That observation by Hanton et al. suggested to us that the dissolved solids were being caused to preferentially crystallize at the droplet-air interface, but not in the core of the droplet, and that subsequent growth of those nuclei in the diffuse layer at the droplet-air interface was favored.
Based on the observations of ion-induced nucleation in solution reported in the previous section, and the clusters that are readily observed in ES-MS, there can exist entities that range in size from macroscopic solids to clusters and nuclei, in the diffuse layer at the droplet air interface of droplets with net charge dependant on the concentration and solubility of the solutes in the starting solution. Do such entities affect the discharge dynamics of Coulomb explosion? To address this question through experiment, we developed a method that allowed the materials separated by Coulomb explosion of a single droplet to be collected at different locations of a target, thereby enabling their differential characterization. In the absence of nuclei or clusters (or at least in the absence of appreciable quantities of nuclei), Cohen et al. measured no difference in the discharge dynamics of droplets generated in an ES when starting solutions consisting of NaCl at 1, 10, and 100 μM were introduced to the ionization source.[152]
The alignment of the ring electrodes of the EDB were changed from perpendicular relative to the target plate, to an angle of 15 degrees relative to the target plate for the following experiments. The electric fields in this configuration of the EDB introduced a force onto the droplets being ejected from the EDB, which caused droplets of different m/z to land on the target at different locations. As a first demonstration of this, droplets were dispensed with a starting solution containing 320 nM fluospheres in 97:3 distilled water:glycerol. One droplet was dispensed at each of the following induction potentials, identified as (i) 100, (ii) 150, and (iii) 200 V dc in
Having established the directionality to the m/z scale of the relative location for the droplet landing on the target, the following experiment was performed. Two starting solutions containing NaCl at 7.5 mM and 37.5 mM in water were used. A 200 V dc potential was applied to the induction electrode. Within experimental error, these droplets from either solution had the same initial volume and net charge at the moment of dispensing. Individual droplets were trapped and levitated in the EDB until they were observed to undergo a rapid oscillation in amplitude, which is characteristic of a droplet undergoing Coulomb explosion.[131] After the main residue had landed on the target, the target plate was translated relative to the ring electrodes along the x-axis for the purpose of isolating main residues between successive iterations of this experiment. The main residues following Coulomb explosion of individual droplets are denoted by the white arrows in
The equation (qe)2=64π2 ∈oγR3 predicts the condition for onset of Coulomb explosion, where γ is the surface tension of a droplet of radius R and having q net elementary charges.[152] According to this equation, the droplets studied here would have shrunk to a radius of ˜9 μm when the explosion occurred, at which point the concentrations of NaCl would have been approximately one order of magnitude below its solubility limit. Hence, we suggest that NaCl precipitates could have existed in the droplets at the time of Coulomb explosion as a result of ion-induced nucleation. It is however known for certain that there were NaCl precipitates in the main residue following Coulomb explosion, and the number of precipitates (in the absence of glycerol) suggests that nucleation initiated at numerous sites. The difference in solute mass could account for the relative deposition locations of the main residues on the target (
For further investigation, another experiment involving two starting solutions consisting of 21 or 320 nM fluospheres in water were used. Fluospheres were used because at similar concentrations, the size of the NaCl(S) would not be decipherable on the target, and the fluospheres are themselves large entities as measured on a molecular scale. Hence the fluospheres are near ideal test solutes to investigate the possibility that the presence of 20 nm diameter entities affect the discharge dynamics of a Coulomb explosion. Droplets were dispensed with one of the two starting solutions loaded into the droplet dispenser and a dc potential of 200 V applied to the induction electrode. The initial volume and net charge of these droplets were the same within experimental error. Droplets were levitated individually until they underwent a Coulomb explosion, as verified through visual observation of the levitated droplet. Following the explosion, the main residues were observed to have been levitated briefly in the EDB prior to falling out of the EDB and onto the target. The target was translated relative to the ring electrodes of the EDB along the direction of the x-axis to permit isolation of each main residue prior to dispensing another droplet. Fluorescence emission from the fluospheres retained in seven main residues are observable in
At the time these droplets underwent a Coulomb explosion, the total mass of fluospheres in any of these droplets was estimated as <2 ppm by mass. Hence, the different concentration of fluospheres in these droplets cannot account for the different m/z of the main residues that are depicted in
The promotion of NaCl nucleation in levitated glycerol droplets was found to be dependent on the magnitude of the net charge in the diffuse layer at the droplet-air interface. This phenomenon is being termed ion-induced nucleation in solution. Some of the potential implications of this result on the performance of ES-MS when characterizing solutes in solutions containing high electrolyte concentrations were discussed. It was speculated that the existence of precipitates and nuclei, such as clusters, in the diffuse layer at the droplet-air interface preceding the explosion influences the discharge dynamics. A first measure of the m/z of the main residues resultant from a Coulomb explosion when the droplets contained 20 nm size entities (fluospheres) indicated that there was a fluosphere concentration dependence on the discharge dynamics.
Several experimental factors regarding precipitation of NaCl in electrodynamically levitated droplets of water/glycerol have been characterized. These droplets carry net elementary charge which is comprised of a population of ions of a single polarity that are in excess of counterions in the droplet. In this Experiment the onset of ion induced precipitation of NaCl was studied as a function of droplet size, surface net excess charge density, viscosity of the droplet, and levitation time at which nuclei were observed to have formed in the levitated droplets as a function of the magnitude of the net charge. The promotion of precipitation of two organic compounds, α-cyano-4-hydroxycinnamic acid (CHCA) and 2,4,6 trihydroxyacetophenone (THAP) was also demonstrated in levitated glycerol droplets as a function of the droplet net charge magnitude. Lastly, investigation of the precipitation of stereoisomers of serine was performed in bulk media that did not have net charge, and in levitated glycerol droplets that possessed net charge. The onset of precipitation for samples of D versus L serine in the levitated droplets of varied net charge was observed to be different. Collectively, these results are examples of effecting a degree of control over the precipitation of dissolved solids. The inventors repeatedly observed in experimental results that indicated the magnitude of the net charge altered the solubility of a dissolved solid, or solids, contained in a levitated droplet. In all cases, decreased solubility for dissolved solids was observed when the magnitude of the net charge on the levitated droplet was increased, likely because of heterogeneous nucleation on or adjacent to an unpaired ion in the levitated droplet.
As discussed above, the inventors have demonstrated ion-induced nucleation to the condensed phase in charged glycerol-water droplets levitated in an electrodynamic balance 12 (EDB). In the case of sodium chloride, ion-induced nucleation was manifested in morphological change, from a cubic shaped crystal to one that contains curved morphology. The promotion of nucleation was onset at a mass-to-charge ratio of 4.78×109 amu/e. While this phenomenon has been observed for several organic and inorganic compounds, the inventors have used sodium chloride as a model solute due to the relative simplicity of the system and the certainty with which the identity of ions carrying the net excess charge (ionsnec) were either Cl− or Na+.
As described above, droplets were dispensed from a starting solution containing the solute of interest and a net charge was imparted on each droplet. The origin of the ions that carried the net charge were either electrolytes added to the starting solution or generated by electrolysis. These ionsnec were localised in a diffuse layer at the droplet-air interface, where they collectively formed an electric potential that diminished to null in the centre of the droplet. Each droplet was captured and levitated in the EDB 12. The volatile solvents evaporated leaving behind a glycerol-water droplet of 11±1 pL that retained the solutes and, in some cases, became supersaturated. Subsequently, droplets were simultaneously deposited on a glass slide and the droplet residues were examined with optical microscopy.
In this Example an experiment was performed to investigate whether the onset of ion-induced nucleation in a droplet is dependent on the surface charge density and hence occurs at the droplet-air interface or on the volume charge density, hence occurring virtually anywhere in the droplet. The inventors varied the volume of the droplets for up to four-folds with corresponding radii 10 μm to 16 μm, while the sodium chloride concentration remained constant and monitored the percent occurrence of ion-induced nucleation as a function of induction potential (
Next, the inventors investigated the effect of the droplet's physical properties on the surface charge density required for ion-induced nucleation to occur, by changing the glycerol/water composition with sodium chloride concentration remaining constant and monitoring the percent occurrence of ion-induced nucleation as a function of induction potential (
The inventors then levitated two populations of water-glycerol droplets containing the same concentration of NaCl, which was low enough not to induce crystal formation at least in the first two hours of levitation. The net charge imparted on both populations was −135 fC and −350 fC, respectively. While the droplets were levitated, the glycerol's slow evaporation gradually reduced the mass-to-charge ratio of the droplets and simultaneously increased the NaCl concentration (
The effect of ion-induced nucleation on the morphology of sodium chloride crystals also depended on the droplet's solvent composition. Curved crystals were more predominant when the water content was ≧10% whereas below that the growth of NaCl crystals was hindered by the high viscosity and the crystals assumed cubic or amorphous shaped and rarely exceeded ˜7 μm in dimension. Below the requisite surface charge density, these crystals were predominantly found in the middle of the droplet's residue which indicates their formation in the core of the droplet, while upon the occurrence of ion-induced nucleation, reproducible homogeneous distribution of small crystal was found on the periphery of the deposited droplet residues in the case of low water content in the droplet.
Although the onset of ion-induced nucleation depends mainly on the droplet's properties, this process had different effects on different solutes. In levitated droplets that contained α-cyano-4-hydroxycinnamic acid (CHCA) and that had a net charge of −135±11 fC, only solids >3.5 μm in dimension were observed (
In the case of THAP, the promotion of nucleation at the surface of the droplet manifested in bundled cylindrical crystals with curved morphology (
Starting solutions containing serine in water were levitated. No glycerol was used since serine seems to act as a surfactant that allows the levitated droplet to retain high enough quantity of water in the EDB in equilibrium with the ambient relative humidity. Hence, by controlling the relative humidity, one could control the final volume of the droplet and, consequently, the concentration of serine in the droplet.
Solutions of D- and L-serine were dispensed and, droplets were levitated at various relative humidities. The droplets were subsequently deposited on a glass slide and the percentage of droplets were the precipitation of serine was recorded (
Pure L- and pure D-serine crystallization occurs at >3M concentration. However L- and D-mixtures of serine has been observed to crystallize at ˜1.5 M. This is a known phenomenon where the crystal building block is composed of both L- and D-serine instead of each alone. This mixed solution could have a potential to precipitate at lower concentration than the pure form. Hence, 9 solutions of D- and L-serine mixtures were prepared whereby the percentage of D- increased from 10% to 90%. Crystallization occurred on all 9 solutions at different rates. Samples were centrifuged, and the supernatant decanted and the peletted precipitates were weighed (
When the same experiment was repeated in levitated droplets, different patterns of crystallization occurred whereby the symmetry of precipitation of serine enantiomers was broken (
Finally, in order to calculate the volume of the droplets and hence the concentration of serine in the droplets as a function of humidity, an experiment was performed on the EDB with glycerol droplets as a standard. 3% glycerol solution was dispensed to create droplet of which we know the mass to charge ratios. The charge was varied and the magnitude of the AC field required to center the droplet in the null point of the EDB was recorded as a function of induction potential (
The following materials and methods were used in this Example.
Starting solutions. The preparation of the starting solutions used to create the droplets for these experiments were as follows: i) 1.1, 2.2, 3.3, or 4.4 mg of NaCl, 4, 8, 12, or 16 μL of glycerol, 396, 392, 388, or 384 μL of distilled deionized water, to create droplets with radii of 10, 12, 14 and 16 μm respectively. For the slow evaporation experiment, the starting solution composition was as follows: (1.1 mg) 28 μmoles of NaCl, 12 μL of glycerol, 388 μL of distilled deionized water.
Droplet dispensing. A 5 μL aliquot of a starting solution was delivered to the reservoir of an ink-jet style droplet generator (model MJ-AB-01-60, Microfab, Piano, Tex., USA). The separation between the induction electrode and the nozzle of the droplet generator was 2 mm. The potential applied to this electrode determined the magnitude of the net charge induced onto each droplet. The net charge on individual droplets were measured by dispensing droplets through a 2.5 mm diameter hole in the induction electrode onto a metal plate connected to an electrometer (model 6517a, Keithley Instruments, Cleveland, Ohio). For induction potentials of 100, 150, and 200 V DC, the induced net charge was −135±11, −235±12, and −325±18 fC respectively. The initial volume of the droplets dispensed was determined by dispensing a starting solution containing 3.7 MBq 32P labeled orthophosphate directly into a liquid scintillation vial. Radionuclide decay was measured using a liquid scintillation counter (LKB Wallac 1217 RackBeta, Fisher Scientific, Montreal, PQ). Droplets had an initial volume of 400±20 μL (average radius 45±2 μm).
Droplet levitation. The droplets passed through the hole cut in the induction electrode and into a double-ring electrodynamic balance (EDB). The volatile solvents evaporated within 5 seconds of their formation, and the volume of the levitated droplets shrank to 11±1 μL (average radius 14 μm). The droplets retained the net charge and the solution became supersaturated. Coulomb explosion 27 was not encountered since the net charge increased linearly with the increase of applied induction potential (unpublished results). Droplets were trapped and levitated in the EDB for 5 minutes, and then deposited onto a plate by applying an attractive DC potential to that plate 28. To determine the condition at which NaCl nucleation in levitated glycerol droplets occurred, each experiment involved the generation and levitation of a population of 30 identical droplets, four of which were deposited at 1 hour intervals for a period of 7 hours in total.
Optical microscopy of droplet residues. A t-test was performed on the diameter of 50 droplet residues measured using a calibrated optical microscope (Motic, B5 Professional, Richmond, BC) for three different magnitudes of droplet net charge; −135 fC, −240 fC, and −350 fC. All droplet residue diameters were similar with 99% confidence level. In 95+% of the droplet levitation trials, crystals were observed to have formed during the period of levitation and no new crystals were observed to have formed in the droplet residue while in contact with the surface of the glass slide up to 4 hours after the time of deposition. In the remaining trials, no crystals were observed to have formed in the droplet residue while it had remained levitated. In these situations, crystals that did form after deposition were a result of heterogeneous nucleation at the interface between the glass slide and the droplet, and their morphology was different than the crystals that formed within the levitated droplet. Similar results were obtained for droplets with net positive charge.
Controlling the size and composition of the droplet. The size of the droplet is controlled through the percent glycerol in the starting solution. In this experiment, starting solutions containing 14% glycerol were used. NaCl concentration in the starting solution was changed accordingly to keep the final solution in the water-glycerol droplet constant. The glycerol-water composition of the droplets was controlled via changing the humidity. Droplets were levitated at 10%, 30%, and 60% relative humidity which corresponds to 3, 10, and 31% water by volume. The viscosities of these solutions were 21, 259, and 800 mPa-sec.
This application claims the benefit of U.S. provisional patent application Ser. No. 60/542,294 filed 9 Feb. 2004 which is hereby incorporated by reference.
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/CA2005/000158 | 2/9/2005 | WO | 00 | 11/24/2008 |
| Number | Date | Country | |
|---|---|---|---|
| 60542294 | Feb 2004 | US |
