DETAILED DESCRIPTION OF THE INVENTION

[0025] The present invention involves protocols for designing and implementing sets of simultaneous experiments, in a parallel, multi-variable process optimization reactor. The multi-variable process optimization reactor is preferably a parallel flow reactor having the operational capability to simultaneously and controllably vary reaction conditions between reaction vessels—either modularly or independently. The simultaneously varied reaction conditions preferably include at least two of the following, in various combinations and permutations: space velocity, contact time, temperature, pressure and feed composition. Compositional variations in the catalysts residing in each of the reaction vessels can also be investigated in the set of simultaneous experiments implemented in the parallel reactor. A preferred multi-variable process optimization reactor is disclosed in co-owned, co-pending U.S. patent application Ser. No. 60/185,566 filed Mar. 7, 2000 by Bergh et al., U.S. Ser. No. 09/801,390 filed Mar. 7, 2001 by Bergh et al., U.S. Ser. No. 09/801,389 filed Mar. 7, 2001 by Bergh et al., and U.S. Ser. No. 09/801,430 filed Mar. 7, 2001 by Srinivasan et al, each of which are incorporated by reference in its entirety for all purposes. Some aspects of this preferred parallel multi-variable optimization reactor are discussed below, in connection with FIGS. 2F through 2T .

[0026] Significantly, as described more fully below, the investigation of various reaction conditions for a reaction of interest in a simultaneous set of parallel experiments (e.g., in a single run through a parallel multi-variable process reactor) provides sufficient data for creating a meaningful master curve of determined selectivity for the reaction of interest versus determined conversion—for each of one or more catalyst compositions being investigated in that single run of parallel experiments. The obtention of sufficient data from a single set of simultaneous experiments to generate a meaningful master curve represents a substantial advance in the art of catalysis research, and particularly, heterogeneous catalysis research.

[0027] The simultaneously varied reaction conditions for the set of simultaneous experiments preferably include a combination of at least two reaction conditions: (i) either different space velocities or different contact times, in combination with (ii) either different temperatures, different pressures or different feed compositions. In some embodiments, the simultaneously varied reaction conditions for the set of simultaneous experiments preferably include different space velocities, in combination with different contact times. More specifically, preferred sets of simultaneously varied reaction conditions include: space velocity and temperature; space velocity and pressure; space velocity and feed composition; contact time and temperature; contact time and pressure; contact time and feed composition; and space velocity and contact time. In other embodiments, higher numbers of reaction conditions, such as three or more reaction conditions or four or more reaction conditions can be simultaneously varied in the set of simultaneous experiments. For example, the simultaneously varied reaction conditions can include one or more of the following preferred tertiary combinations of reaction conditions: different space velocities, temperatures and pressures; different space velocities, temperatures and feed compositions; different space velocities, pressures and feed compositions; and different temperatures, pressures and feed compositions. A preferred approach for simultaneously varying four different reaction conditions can include a combination of either: different space velocities, temperatures, pressures and feed compositions, or different contact times, temperatures, pressures and feed compositions.

[0028] Generally, the terms used herein to describe the various reaction conditions, including temperature, pressure, space velocity, contact time and feed composition have their ordinary meaning as understood by a person of ordinary skill in the art. Temperature and pressure are independent variables, whereas space velocity, contact time, and feed composition are partially coupled variables (i.e., two of the three can be independently controlled). Briefly, contact time is a measure of the total average time of exposure between a fluid and a catalyst in a flow reactor, and is typically defined as the volume of the reaction zone divided by the total volumetric flowrate of the feed. Space velocity is a measure of the molar, mass or volume flowrate of the reactant(s) or key reactant—and an be controlled independently of the total molar flowrate of feed (e.g., due to differences in concentration of the reactant(s) in the feed, such as due to dilution of reactants with one or more inert fluids). Feed composition refers to a measure of the relative ratio of various reactant components and/or inert components of the feed mixture. A more detailed discussion of these reaction condition parameters is provided in many texts. See, for example, Wijngaarden et al., “Industrial Catalysts—Optimizing Catalysts and Processes”, Wiley-VCH, Germany (1998).

[0029] In preferred protocols for any of the aforementioned embodiments, at two different space velocities, and preferably at least three or at least four or at least five or at least six different space velocities are simultaneously investigated in combinations with the other reaction condition(s) as described. Similarly, at least two, preferably at least three or at least four or at least five or at least six different temperatures are simultaneously investigated. Likewise, other reaction conditions (pressures, feed compositions, contact times) are simultaneously investigated using at least two, preferably at least three or at least four or at least five or at least six different values of such other reaction conditions. Particularly preferred combinations of reaction conditions to be simultaneously evaluated in a single set of parallel experiments include at least three space velocities with at least two temperatures, pressures, feed compositions or contact times. In some embodiments, at least six space velocities are evaluated simultaneously with at least four temperatures, pressures, feed compositions or contact times.

[0030] Generally, with reference to FIG. 2A, a method for evaluating process conditions for a catalyzed chemical reaction in a parallel flow reactor comprises simultaneously supplying one or more reactants through a fluid distribution system to each of six or more reactors of a parallel flow reactor under reaction conditions to effect a chemical reaction of interest, while controllably varying one or more sets of reaction conditions between each of the six or more reactors. A reactor effluent containing one or more reaction products and, in some cases, one or more unreacted reactants is simultaneously discharged from each of the six or more reactors. Each of the reactor effluent streams can be sampled and analyzed, preferably simultaneously analyzed, to determine the conversion of one or more of the reactants, and the selectivity for at least one reaction product for the reaction of interest. The conversion is preferably determined for the conversion-limiting reactant. Generally, as used herein, conversion refers to a measure of the extent to which a reaction has proceeded, and is typically defined as the fractional amount of the amount of reactant that has been converted to one or more products (regardless of whether the products are of interest, or are side reaction products). The selectivity refers to a measure of the extent to which a reaction resulted in a desired product of interest, and is typically defined as a ratio of the amount (e.g., molar amount) of desired product obtained to the amount (e.g., molar amount) of the key reactant converted. A more detailed discussion of conversion and selectivity is provided in Wijngaarden et al., “Industrial Catalysts—Optimizing Catalysts and Processes”, Wiley-VCH, Germany (1998).

[0031] Each of the six or more reactors comprises a catalyst having activity for the chemical reaction of interest. In general, depending on the number of reaction vessels in the parallel reactor, at least two or more, preferably at least four or more, and in some embodiments at least six or more of the catalysts are substantially the same—such that they have substantially the same composition and/or were prepared by substantially the same synthesis protocols with substantially the same compositional recipe and/or were prepared by substantially the same mechanical (e.g., grinding, pressing, crushing, sieving) treatments, chemical treatments, and/or physical treatments.

[0032] In preferred embodiments, the one or more sets of reaction conditions are controllably varied such that a determined conversion (e.g., the conversion of the conversion-limiting reactant) and a determined selectivity for one or more reaction products of the reaction of interest includes at least two, preferably at least three, preferably at least four, more preferably at least five and most preferably at least six data values for each of the catalyst compositions being evaluated. The determined data values for conversion preferably span a range of values that relate to the conversion range of interest for the chemical reaction, such that a meaningful master curve can be generated for each of the catalyst compositions being evaluated. Generally, the determined conversion values include four or more values, preferably six or more values, that span a range of at least about 10% conversion difference, and more preferably at least about 20% conversion difference between the highest and lowest of such values, and for many reactions of interest, more preferably at least about 30%, 40%, 50% or 60% conversion difference between the highest and lowest of such values. Considered in another manner, the determined conversion values (e.g., of the conversion-limiting reactant) include four or more values, preferably six or more values ranging from less than about 20% conversion to more than about 40% conversion. That is, within the determined three or six values for conversion, at least one of the determined values is less than about 20% conversion, and at least one of the determined values is more than about 40% conversion. Preferably, the set of reaction conditions are varied in the single set of simultaneous parallel reactions such that the determined conversion values (e.g. of the conversion-limiting reactant) include a range of four or more values, preferably six or more values, ranging from less than about 15% conversion to more than about 45% conversion, preferably ranging from less than about 10% conversion to more than about 50% conversion, and in some embodiments, ranging from less than about 10% conversion to more than about 70% conversion, or even ranging from less than about 10% conversion to more than about 80% conversion. For some reactions of interest (e.g., propylene oxidation to propylene oxide; e.g., benzene to aniline conversions), relatively lower conversions are commercially significant due to economics, thermodynamics and/or safety considerations, typically with relatively higher selectivities. Hence, for some reactions of interest, the determined conversion values (e.g., of the conversion-limiting reactant) include three or more values, preferably four or more values, more preferably six or more values ranging from about 2% or less to more than about 5%, preferably from about 1% or less to more than about 10% conversion, more preferably from about 1% or less to more than about 15%. That is, within the determined three or six values for conversion, at least one of the determined values is about 2% or less (preferably about 1% or less), and at least one of the determined values is more than about 5% conversion (preferably more than about 10%, more preferably more than about 15% conversion).

[0033] Preferred Evaluation Protocols

[0034] In one preferred embodiment, a parallel reactor having six or more reaction vessels is loaded with a set of six or more of substantially the same catalysts (e.g. having substantially the same composition). A set of reaction conditions is controllably varied between the six or more reaction vessels, with the particular combinations of varied reaction conditions being selected generally from those described above. Preferably, with reference to FIG. 2 B, the set of reaction conditions varied between the six or more reactors (having substantially the same catalyst, indicated as “C 1 ” in FIG. 2B ) include (i) at least three different space velocities (SV 1 , SV 2 , SV 3 ) or alternatively, at least three different contact times (CT 1 , CT 2 , CT 3 ), and (ii) at least two different temperatures (T 1 , T 2 ), or alternatively, at least two different pressures (P 1 , P 2 ), or alternatively, at least two different feed compositions (FC 1 , FC 2 ). As shown in FIG. 2 B, this set of parallel experimental conditions can be represented schematically in 3×2 matrix with the catalyst (C 1 ) indicated in each box (e.g., such that each box represents a separate reaction vessel of the parallel flow reactor) and with the reaction conditions applied to the catalyst-containing reactors being indicated along each side of the matrix.

[0035] In a variation of the immediately preceding embodiment, a parallel reactor having twelve or more reaction vessels is loaded with a set of twelve or more of substantially the same catalysts (e.g. having substantially the same composition). Two sets of reaction conditions are controllably, and independently varied between six or more reaction vessels per set—a first set of reaction conditions being varied between a first set of six or more reactors, and a second set of reaction conditions being varied between a second set of six or more reactors, with the particular combinations of varied reaction conditions being selected generally from those described above. Preferably, with reference to FIG. 2 C, the first set of reaction conditions varied between the first set of six or more reactors (e.g., having catalyst C 1 ) include (i) at least three different space velocities (SV 1 , SV 2 , SV 3 ) or alternatively, at least three different contact times (CT 1 , CT 2 , CT 3 ), and (ii) at least two different temperatures (T 1 , T 2 ). The second set of reaction conditions varied between the second set of six or more reactors (e.g., having catalyst C 1 ) include (i) at least three different space velocities (SV 1 , SV 2 , SV 3 ) or alternatively, at least three different contact times (CT 1 , CT 2 , CT 3 ), and (ii) at least two different pressures (P 1 , P 2 ), or alternatively, at least two different feed compositions (FC 1 , FC 2 ). The space velocity and/or contact time can be the same or different as compared between the first set of reaction vessels and the second set of reaction vessels. As shown in FIG. 2 C, this set of parallel experimental conditions can be represented schematically in 3×4 matrix.

[0036] In another preferred embodiment, a parallel reactor having twelve or more reaction vessels is loaded with twelve or more catalysts—each having activity for the chemical reaction of interest. A first set of the twelve or more catalysts can be six or more first catalysts that are substantially the same (e.g. having substantially the same first composition). A second set of the twelve or more catalysts can be six or more second catalysts that are substantially the same (e.g. having substantially the same second composition). A first set of reaction conditions is controllably varied between the reaction vessels comprising the first set of catalysts, with the particular combinations of varied reaction conditions being selected generally from those described above. Similarly, and simultaneously therewith, a second set of reaction conditions is controllably varied between the reaction vessels comprising the second set of catalysts, with the particular combinations of varied reaction conditions being selected generally from those described above. Preferably, with reference to FIG. 2 D, the first set of reaction conditions, and the second set of reaction conditions are varied between their respective reaction vessels to include (i) at least three different space velocities (SV 1 , SV 2 , SV 3 ) or alternatively, at least three different contact times (CT 1 , CT 2 , CT 3 ), and (ii) at least two different temperatures (T 1 , T 2 ), or alternatively, at least two different pressures (P 1 , P 2 ), or alternatively, at least two different feed compositions (FC 1 , FC 2 ). As shown in FIG. 2 D, this set of parallel experimental conditions can be represented schematically in 3×4 matrix.

[0037] In a further preferred embodiment, a parallel reactor having sixteen or more reaction vessels is loaded with sixteen or more catalysts—each having activity for the chemical reaction of interest. The sixteen or more catalysts include at least four sets of catalysts, each set having four or more catalysts. The catalysts within each of the four sets are substantially the same (e.g., or at least having substantially the same composition), but the catalysts as compared between different sets vary with respect to each other. Specifically, the sixteen or more reactors can comprise a first set of four or more of the catalysts having substantially the same first composition, a second set of four or more of the catalysts having substantially the same second composition, a third set of four or more of the catalysts having substantially the same third composition, and a fourth set of four or more of the catalysts having substantially the same fourth composition. A set of reaction conditions is controllably varied between the sixteen or more reaction vessels—with the particular combinations of varied reaction conditions being selected generally from those described above. Preferably, with reference to FIG. 2 E, the set of reaction conditions are varied between to include (i) at least four different space velocities (SV 1 , SV 2 , SV 3 , SV 4 ) or alternatively, at least four different contact times (CT 1 , CT 2 , CT 3 , CT 4 ), and (ii) at least four different temperatures (T 1 , T 2 , T 3 , T 4 ), or alternatively, at least four different pressures (P 1 , P 2 , P 3 , P 4 ), or alternatively, at least four different feed compositions (FC 1 , FC 2 , FC 3 , FC 4 ). Additionally, the set of reaction conditions are varied such that at least one catalyst from each of the first, second, third and fourth sets of catalysts sees each of the reaction conditions (considered independently, such that not all catalysts see all of the possible combinations of conditions). That is, the set of reaction conditions are varied such that at least one catalyst from each of the first, second, third and fourth sets of catalysts that catalyzes the chemical reaction under each of the at least four different space velocities or contact times, and under each of the at least four different temperatures, pressures or feed compositions. As shown in FIG. 2 E, this set of parallel experimental conditions can be represented schematically in 4×4 matrix. In a variation of this approach, the number of reaction conditions being varied can increase for evaluation of the four or more sets of different catalysts. Specifically, for example, the parallel flow reactor can comprise twenty-four or more reactors to effect the chemical reaction of interest, with each of the twenty-four or more reactors comprising a catalyst having activity for the chemical reaction. The varied set of reaction conditions can comprise (i) at least six different space velocities or contact times, and (ii) at least four different temperatures, pressures or feed compositions, and the set of reaction conditions can be varied such that at least one catalyst from each of the first, second, third and fourth sets of catalysts catalyzes the chemical reaction under each of the at least six different space velocities or contact times, and under each of the at least four different temperatures, pressures or feed compositions, although not necessarily under each permutation of combinations of such reaction conditions.

[0038] Advantageously, the immediately preceding experimental protocol for a multi-channel, multi-variable set of simultaneous experiments provides sufficient data to generate a meaningful master curve, without performing as many experiments as would be necessary to provide data for each of the four or more compositions at each particular combination of reaction conditions. The utility of this approach can be enhanced when combined with one or more interpolation and/or extrapolation techniques, such as predictive extrapolation based on a defined similarities with a related reference composition. To this end, in one variation of the embodiment described immediately above, the parallel flow reactor can comprises twenty or more, preferably twenty-four or more reactors of a parallel flow reactor to effect the chemical reaction of interest, each of the twenty or more, preferably twenty-four or more reactors comprising a catalyst having activity for the chemical reaction of interest. Four or more, and preferably eight or more of the reactors can comprise a reference set of four or more, preferably eight or more reference catalysts having substantially the same reference composition. The reference composition can be a standard composition that is a representative composition for the four or more sets of different catalysts (C 1 , C 2 , C 3 , C 4 ) being evaluated. As a non-limiting example, the reference composition can be the same catalyst platform and have the same major components (e.g., components having a relative molar ratio of more than about 10%) as one or more of the other catalysts being evaluated, but different minor components (e.g. dopants) having a relative molar ratio of 10% or less). As another non-limiting alternative, the reference composition can be the same catalyst platform and have the same or functionally similar components with a relative molar ratio of components that is about the average of the range of molar ratios being evaluated in the library. In any case, the reference composition is repeated at least four times, preferably at least eight times, and the reaction conditions to which the reference compositions are exposed are preferably varied in the same manner as they are varied for the four or more sets of different catalysts, such that the four or more, preferably eight or more see at least four or more, preferably eight or more of the varied reaction conditions or combinations thereof. For at least one, and preferably for each of, the first, second, third or fourth catalyst compositions, at least a portion of the master curve can be interpolated or extrapolated by comparison with a master curve determined for the eight or more reference catalysts. The comparative interpolation and/or extrapolation can be visual or qualitative, and/or can be based on mathematical and statistical modeling approaches known in the art. In addition to interpolation and/or extrapolation between determined data points based on a master curve for the reference catalyst, one can also interpolate and/or extrapolate entire master curves (or substantial portions thereof) based on comparison to master curves developed for similar catalysts using one or more of the approaches described herein. Further details about such a predictive approach are described below in connection with FIGS. 5A through 5E .

[0039] In any of the aforementioned embodiments, additional blank reaction channels—having an essential absence of catalytic activity for the reaction of interest—can also be employed, for example, to determine background and/or to detect reaction conditions that may decompose reactants or products.

[0040] In yet a further application, a parallel process optimization reactor can be used to effect a single set of simultaneous (parallel) experiments to check for diffusion limitations, useful for example in screens directed toward determining intrinsic activity and/or kinetic activity of a catalyst composition. Briefly, a parallel reactor having twelve or more reaction vessels is loaded with a set of twelve or more of substantially the same catalysts (e.g. having substantially the same composition). Two sets of reaction conditions are controllably, and independently varied between six or more reaction vessels per set—a first set of reaction conditions being varied between a first set of six or more reactors, and a second set of reaction conditions being varied between a second set of six or more reactors. Specifically, the varied first set of reaction conditions generally comprises varied average particle size of the catalysis materials, prepared, for example, as described in co-owned, co-pending, U.S. patent application Ser. No. ______ entitled “Methods and Apparatus for Mechanical Treatment of Materials Such as Catalysts” filed on the date even herewith by Lugmair et al. The varied second set of reaction conditions can include linear gas flowrates through a reaction zone (e.g., through a fixed bed of a fixed bed reactor). Preferably, the varied first and second sets of reaction conditions can be employed in conjunction with a substantially constant space velocity and/or contact time. For example, increased gas flow can be realized without substantial effect on space velocity and/or contact time if the catalyst mass and bed height are also increased, or if a diluent (e.g., SiC) is employed, by varying the amount of the diluent. By varying particle size in six or more reaction vessels, and by varying linear (gas) reactant flowrate in six or more vessels, one can evaluate whether film diffusion and/or pore diffusion are limiting. The diffusion-limited nature of a reaction (or the lack of a diffusion-limited nature) is helpful in evaluating intrinsic activity or kinetic activity.

[0041] As noted, although described herein in connection with heterogeneous catalysis research with a parallel flow reaction system, the present invention has applications to homogeneous reaction systems, and for both heterogeneous and homogeneous catalysts, to parallel semi-continuous reaction systems or to parallel batch reaction systems. Furthermore, it is contemplated that other specific variations and combinations of parameters in a multidimensional space can be employed. See, for example, U.S. patent application Ser. No. 60/198,208 entitled “Combinatorial Parameter Space Experiment Design” filed Apr. 19, 2000 by Falcioni et al, which is hereby incorporated by reference in its entirety for all purposes.

[0042] Additional details about the preferred reaction system, about combinatorial catalysis research approaches, about preferred candidate materials and libraries of candidate materials, about preferred reactions of interest, and about various alternative approaches for evaluating catalysts—for use in connection with the above-described protocols—are described below.

[0043] Parallel Multi-Variable Process Optimization Reactor

[0044] A preferred parallel, multi-variable optimization reactor suitable for use in connection with the methodologies of the invention is shown in FIGS. 2F through 2T . With reference to FIG. 2 F, the twenty-four channel reaction system 1000 is a bench-top instrument comprising a distribution module 4500 and a reactor module 4600 . The distribution module 4500 and reactor module 4600 are supported in a frame 4100 that includes a lower support shelf 4110, and upper support shelf 4130 and four guide posts 4120 (linear ball bearings). The distribution module 4500 can be operationally engaged with the reactor module 4600 by downward movement of a shaft 4142 connected at a lower end to the distribution module 4500 via a support block (not shown), and at an opposing upper end to a hydraulic jack 4140 (e.g. 100 kN, adjustable, regulated force) supported on a press frame 4145 . Preloaded springs 4146 are provided on each of the guide posts 4120 to ensure appropriate seating force between the reactor module 4600 and the lower support shelf 4110 to sequentially demount the distribution module from the reactor module (release of upper o-ring seals on reactor module), and subsequently, the reactor module lower o-rings). A gas-chromatograph-connector module 4200 provides twenty-four channel fluidic interface to a parallel gas chromatograph injection valve (not shown) situated under the lower support shelf 4110 , and is in fluid communication with the reactor module 4600 .

[0045] Referring to FIG. 11 G, the distribution module 4500 can be interfaced with an external fluid distribution system 1480 having gas supply system 1481 with associated mass flow controllers 1483 , and a liquid supply system 1482 with associated evaporators 1484 , together with appropriate valving, substantially as described above. Three gaseous feed components are fed through external transfer lines 1530 ′ (variable partial pressure reactant feed source), 1532 ′ and (constant partial pressure reactant feed source), 1534 ′ (make-up gas) substantially as shown to an internal inlet fluid distribution system 1500 having an internal feed-composition subsystem 4525 . Internal feed transfer lines 1530 , 1534 , and pairs of commonly-actuated inlet isolation valves 1487 , 1489 provide selectable fluid communication to a series of six sets of flow restrictor groups—indicated as SET A, SET B, SET C, SET D, SET E, and SET F—with each set comprising six groups of flow restrictors. Each group of flow restrictors in a particular set include a first-feed-component flow restrictor in selectable fluid communication with the variable partial pressure feed source through transfer line 1530 and inlet isolation valve 1487 , and a second-feed-component flow restrictor in fluid communication with the make-up feed source through transfer line 1534 and inlet isolation valve 1489 . The relative conductance values for the first-feed-component flow restrictor and the second-feed-component flow restrictor within each group is indicated as “C 1 ”, “C 2 ”, “C 4 ”, “C 6 ”, “C 8 ” and “C 9 ”, with combinations of values rotating between the various groups as appropriate. In one configuration, internal feed transfer line 1532 provides a constant partial pressure feed source through a set of six dedicated flow restrictors, where the flow restrictors have substantially the same resistance to flow (i.e., “constant resistance” flow restrictors, indicated as SET CR). The (varied) feed composition from the six groups of feed-component flow restrictors (after mixing feed components from the first-feed-component flow restrictors and second-feed-component flow restrictors in an internal mixing zone, not shown,) are fed through six respective discharge channels 4540 . The combined feed composition (of the variable feed component/make-up feed component) are mixed with the constant feed component in the discharge channels 4540 . A pressure sensor (illustrated as p 1 , p 2 , p 3 . etc.) monitors discharge channel 4540 pressure. Each of the resulting six varied feed composition feeds are discharged through an outlet isolation valve 4580 , and then split four ways using flow-splitters to create twenty-four streams—with six groups of four steams each, and each group having a varied feed composition. Each of the twenty-four feed streams can be fed through a single, individual split restrictor 4570 —such as a capillary or as microfluidic channel (e.g. having substantially the same resistance to flow as compared between reaction channels), and then through a mass flow sensor 4590 (MFS), before leaving the distribution module. As illustrated, the order of the mass flow sensor 4590 and the split restrictor 4570 can be reversed from that described above. The distribution module 4500 is a separate, stand-alone subsystem that mounts in fluid communication with the reactor module 4600 . Connector module 4200 and detection module 10000 are also shown.

[0046] The distribution module 4500 was provided as a modular unit comprising a plurality of modular fluidic chips. In particular, with reference to FIG. 2 H, the distribution module comprises a flow-restrictor block 4510 comprising a substrate 3600 having several microchip bodies (not shown) mounted thereon. The microchip bodies include flow restrictors (C 1 , C 2 , C 3 , etc.) from which sets of flow-restrictor groups are formed by fluidic connection. A series of commonly-actuated groups of microvalves can be precision machined and mounted as a valve block 4487 on one side of the distribution module 4500 , or integrally therewith, including for example as a subblock integrally within the flow-restrictor block 4510 —for example, as indicated as 4587 ′ with dotted lines. Alternatively, the groups of microvalves can be microfabricated and be integral with the flow restrictor block 4510 or with microchip bodies mounted on the flow-restrictor block 4510 . The microvalves are pneumatically actuated with air supplied through control pressure ports 4512 . Reactant sources, including a variable partial pressure gas, a make-up gas and a constant partial pressure gas are supplied via external-internal inlet ports 4530 , 4534 , 4532 , to provide fluid connection with internal transfer lines 1530 , 1534 , 1532 , respectively ( FIG. 2G ). A cover 4504 and an insulator block 4502 are positioned over the flow-restrictor block 4510 . The flow-restrictor block 4510 is situated over a heater block 4520 comprising resistive cartridge heaters 4522 , pressure-sensor ports 4524 and thermocouple ports 4526 . Capillary-type split restrictors ( 4570 , FIG. 2G ) extend between the heater block 4520 and the split-restrictor-mass flow sensor block 4550 (“SR-MFS block” 4550 ), that includes an array of microfabricated mass flow sensors 4590 downstream from each of the split restrictors, and connection ports 4552 , 4553 for cooling fluid circulating through the SR-MFS block 4550 . The feed passes from the mass flow sensors into reactor tubes (not shown) that extend upward from the reactor module ( 4600 , FIG. 2G ) into a press block 4560 when the distribution module 4500 is engaged with the reactor module.

[0047] The flow restrictor block 4510 includes six sets of flow-restrictor-groups microfabricated on separate microchip bodies 3650 that are mounted on a common substrate 3600 . In one embodiment, a set of seven microchip bodies having integral flow restrictors can be mounted on a substrate, with each the flow restrictors in each of the mircochip bodies corresponding to one of the sets of flow-restrictor groups (SET A, SET B, SET C, SET D, SET E, SET F, SET CR) represented in FIG. 2G . Exemplary microchip bodies, corresponding to SET A and SET CR are shown in FIG. 21 and 2 J, respectively. The microchip bodies comprise a first inlets 5002 (e.g. in fluid communication with one of the feed-component source gasses, such as the variable-feed component), and a second inlet 5004 (e.g., in fluid communication with another of the feed-component source gas, such as the make-up feed component). When a particular set of flow restrictors is selected by actuation of microvalves ( 1487 , 1489 , FIG. 2G ) corresponding to that set, then the two feed components come in through the inlets 5002 , 5004 into the inlet plenums 5003 , 5005 , and flow through the flow restrictors (e.g., for SET A, varied flow restrictors C 1 , C 2 , C 4 , C 6 , C 8 , C 9 ; for SET CR, constant restrictors CR) to mixing zones 1540 , such that the varied feed compositions having various ratios of feed components are formed in the mixing zones. Microfluidic outlets 5010 provide fluid communcation to the discharge (channels 4540 , FIG. 2G ). As an alternative, geometry effects associated with variously-sized channels off of the common inlet plunums 5003 , 5005 (e.g. entrance volume effects) can be minimized physically forming, for example, the C 9 flow restrictor from nine multiple, identical copies of the C 1 flow restrictor. With reference to FIG. 2 K, in a further embodiment, the flow resistances on each particular microchip body be substantially the same (e.g., C 4 , as illustrated, for one microchip body, C 6 for another microchip body, etc.). Also, each flow restrictor (e.g., C 4 , as illustrated) can have its own, dedicated micrfluidic inlet 5002 and inlet plenum 5003 . Referring to FIG. 2 L, the sets of flow restrictors (with grouped resistance values as shown in FIG. 2G ) can be established where each of the microchip bodies are fabricated to be identical (e.g., such as shown in FIG. 21 )—without physically and integrally rotating the various combinations of flow resistances on the microchip body—by having microfluidic channels 4540 that cross the outlets 5010 of each flow restrictor (C 1 , C 2 , C 4 , etc.) in the various combinations, so that by selection of appropriate feed components to the inlets 5002 , the desired sets of flow restrictors (e.g. SET A, SET B, SETC, etc.) can be achieved. Internal passages within the flow-restrictor block 4520 and/or cover block 4510 , 4504 can be used for internal interconnections. Such internal passages can be supplemented by external interconnections that interface through side ports 4544 . Alternatively, such side-ports could be manifolded and routed through a different face of the flow-restrictor block 4510 . Fabrication of the flow restrictors integral with the microchip bodies can be effected, for any of the aforementioned embodiments, using typical microfabrication techniques.

[0048] The flow restrictor block 4510 also includes the six pairs of commonly-actuated inlet isolation microvalves 1487 , 1489 , as well as the outlet isolation valves 4580 . These valves are preferably fabricated using precision machining techniques known in the art. Alternatively, the valves can be microfabricated, and can be integral with the flow-restrictor block 4510 or with a microchip body mounted thereon. The valves can also be, as noted above, part of an external fluid distribution system ( 1480 , FIG. 2G ). The particular microvalve design is not critical. Preferably, the microvalves 1487 , 1489 are membrane-actuated, membrane-seated valves such as shown in FIGS. 2M and 2N . Briefly, membrane-actuated valves 4300 can be prepared by precision machining to form the various component parts. In its open state ( FIG. 2M ), a fluid can flow into the valve through fluid inlet passage 4302 , through internal passages 4303 , past the valve seat 4310 , and out through outlet passage 4304 . In its closed state ( FIG. 2N ), a piston 4320 having a piston face 4322 is forced upward against a seating membrane 4315 such that fluid flow past the seat 4310 is sealingly blocked, with the seating membrane 4315 essentially acting as a gasket between the piston face 4322 and the valve seat 4310 . The piston 4320 is preferably pneumatically actuated by use of an actuating membrane 4325 under pressure through actuation passage 4330 . Portions of the seating membrane 4315 and actuating membrane that are situated between facing component surfaces of the valve body can serve as gaskets when the valve is clamped or fastened together. Further details are provided in co-pending, co-owned application U.S. Ser. No. 60/274,022 entitled “Gas Chromatograph Injection Valve Having Microvalve Array” filed on Mar. 7, 2001 by Bergh et al.

[0049] The mass-flow-sensor/split restrictor block 4550 is shown in FIGS. 20 and 2 P. The MFS-SR block 4550 comprises, for each of the twenty-four reaction channels, a split restrictor 4570 , and a microfabricated mass-flow sensor 4590 . As shown in FIG. 20 , the unit is cooled using fluid-type micro heat exchanger with cooling air as the cooling medium. The split restrictor is preferably a capillary-type flow restrictor (e.g., {fraction (1/16)} O.D. /125 μm I.D. stainless steel capillaries). For each reaction channel, the feed gas flows through the split restrictor into the mass-flow sensor 4590 via sensor inlet 4591 . The feed gasses passes through a detection channel 4592 and then exits via sensor outlet 4593 . O-rings 4599 are used to seal the sensor inlet and outlet. 4591 , 4593 . The mass-flow sensor design is not critical. Referring, briefly, to FIGS. 2Q, 2R , 2 S, a preferred microfabricated mass flow sensor 4590 includes five-detection filaments 4595 . The detection filaments 4595 are electrically connected to contact pads 4596 via conductive paths. The contact pads are connected to mass-sensor electronic circuitry for a five-bridge design, according to known techniques. The detection filaments 4595 are preferably platinum vapor-deposited onto a silicon nitride bridge, and designed with a meandering path ( FIG. 2S ). The silicon nitride bridge is positioned at 15° angle relative to the normal to the detection channel 4592 . The electronic circuitry (e.g. printed circuit board) can be located adjacent each mass flow sensor (e.g., in the adjacent cavity 4598 ). After exiting the mass flow sensor 4590 , the feed gas for each channel is fed to the reactor tubes 4610 .

[0050] The reactor module 4600 , shown schematically in FIG. 2 T, comprises a 4×6 array of twenty-four reactor tubes 4610 individually supported in a reactor frame 4605 . Each tube has a reaction volume of about 1 ml. Each of the reactor tubes 4610 can be individually heated using resistive coil heaters 4620 (e.g. Watlow Mini-K-ring). Thermal isolation between reactor tubes 4610 is achieved using fluid-type heat exchanger to cool the inter-reactor volume within the reactor frame 4610 . Preferably, the cooling medium is air or inert gas, and is fed into the reactor