Molecular Recognition of Organic Vapors by Adamantylcalix[4]arene

Nov 14, 2008 - Materials and Transducers Toward Selective Wireless Gas Sensing. Radislav A. Potyrailo , Cheryl Surman , Nandini Nagraj , and Andrew ...
0 downloads 0 Views 427KB Size
Download PDF
J. Phys. Chem. B 2008, 112, 15569–15575

15569

Molecular Recognition of Organic Vapors by Adamantylcalix[4]arene in QCM Sensor Using Partial Binding Reversibility Luidmila S. Yakimova,† Marat A. Ziganshin,† Vladimir A. Sidorov,§ Vladimir V. Kovalev,‡ Elvira A. Shokova,‡ Viktor A. Tafeenko,‡ and Valery V. Gorbatchuk*,† Institute of Chemistry, Kazan State UniVersity, KremleVskaya 18, Kazan 420008, Russia; Department of Chemistry, Virginia Commonwealth UniVersity, Richmond, Virginia 23284; and Department of Chemistry, Moscow State UniVersity, Moscow 119992, Russia ReceiVed: May 14, 2008; ReVised Manuscript ReceiVed: September 29, 2008

The parameters of stability, guest binding reversibility, and Gibbs energy of guest inclusion were determined for clathrates of adamantylcalix[4]arene (1). These data provide a new insight into the structure-property relationships in vapor sensor applications of clathrate-forming hosts. A thin layer of 1, used in the quartz microbalance (QCM) sensor, demonstrates a selectivity for organic vapors, which depends on the regeneration technique after the guest binding. Complete regeneration of 1 on the sensor surface was reached through the exchange of bound guest with ethanol vapor, which forms an unstable clathrate with 1. The efficiency of the used regeneration technique was proved by comparing the QCM data with the isotherms of guest vapor sorption by guest-free host 1 and with the data of simultaneous thermogravimetry and differential scanning calorimetry for the saturated clathrates of 1. In sensor, the extent of host regeneration without guest exchange depends on the guest molecular structure. This extent, or guest-binding reversibility parameter, being determined in a combination with the sensor responses of completely regenerated 1 to guest vapors, increases the recognition capability of single sensor device. Using this technique, 13 of 15 studied guests were discriminated. The structural hints on the suitable sensor properties of 1 were found in the determined X-ray monocrystal data for clathrate of this host with toluene. Introduction Solid calixarenes are perspective materials for molecular recognition of organic vapors due to their high binding selectivity.1,2 The main factor contributing to this selectivity is the cooperativity of host-guest clathrate formation. This cooperativity is observed through the change in packing of host molecules in the crystal3–5 and in the sigmoidal shape of the guest sorption isotherms, which display a relative vapor pressure (activity) threshold in guest binding.1,2 Small variations in the guest molecular structure can shift the binding threshold up to a unity, offering almost absolute discrimination between two close guest homologues.2 This property is often desired for a receptor to be used in sensors. Still, the usefulness of calixarenes in vapor sensors can be limited by the low reversibility of guest binding, which is a back side of the guest binding cooperativity in clathrates. A good negative example is the negligibly small response of the tert-butylcalix[4]arene-based quartz crystal microbalance (QCM) sensor for the guest vapors,6,7 comparing with guest/host ratios in saturated clathrates.8 The small value of the response is probably due to the high kinetic and thermal stability of tertbutylcalix[4]arene clathrates.1,2,5,9,10 Such high stability of clathrates precludes the sensor regeneration in a reasonable time. A conceptual change of the substituent pattern in calixarene host may provide a solution to this problem.11 We reasoned that adamantylcalix[4]arene (1), a host that features an extended, * Corresponding author: e-mail [email protected]; Ph +7-8432315309; Fax +7-843-2927418. † Kazan State University. ‡ Virginia Commonwealth University. § Moscow State University.

compared to tert-butylcalix[4]arene, cavity may provide an additional space for guest inclusion and may demonstrate a different clathrate stability and guest binding reversibility, respectively. The X-ray structure of 1 clathrate with toluene, determined in the present work, is in line with this hypothesis. Being a complex function of guest and host structure, the parameter of host-guest binding reversibility may be used for molecular recognition of guest vapors in addition to a sensor response for their binding. To estimate this reversibility in sensors, the guest-free host layer on the sensor surface has to be prepared. In this study, the ability of host-guest clathrates to exchange the guests12,13 was used to prepare such guest-free host layer. The guests, which could not be removed from their clathrates with 1 within the temperature limits of the sensor heating, were exchanged for ethanol. The ethanol clathrate was further completely decomposed by purging with hot air. A ratio of sensor responses for host layer regenerated with and without guest exchange was taken as the value of host-guest binding reversibility parameter. To prove the complete regeneration of host 1 through the guest exchange and to reveal the specific features of its behavior in sensors, the clathrate stoichiometries calculated from the QCM sensor response were compared with those determined in experiments where the absence of guest in host 1 powder was determined directly. For this, isotherms of guest vapor sorption were determined using headspace gas chromatographic analysis for solid 1, previously equilibrated at high temperature (220 °C) to remove possible memory effects. The vapor sorption experiment gives the data on the guest-host binding capacity, cooperativity, and affinity for the initially guest-free host,1,2 and models the host-guest binding in sensors. The composition and stability of clathrates obtained in this vapor sorption process

10.1021/jp804277u CCC: $40.75  2008 American Chemical Society Published on Web 11/14/2008

15570 J. Phys. Chem. B, Vol. 112, No. 49, 2008

were studied by simultaneous thermogravimetry and differential scanning calorimetry (TG-DSC). Once estimated, the binding reversibility, in addition to a response of completely regenerated sensor to guest vapors, gives two quantitative parameters determined for one guest-host pair. In the present work, these two values were found to have different relationships with guest structure. Hence, more effective recognition of guest vapors was performed using a single sensor. Experimental Methods Materials. Adamantylcalix[4]arene (1) was synthesized according to a previously described procedure14 and purified from nonvolatile impurities by multiple recrystallization. The volatile impurities were removed by heating for 8 h at 220 °C in a vacuum (100 Pa). This gives thermally stable form of host 1 (R-phase). The host purity was confirmed by methods described elsewhere.15,16 The content of the host main substance determined by TLC was at least 99%. The purity of guests checked by GC was at least 99.5%. Determination of Vapor Sorption Isotherms. The vapor sorption isotherms were determined using the static method of headspace GC analysis.15 In a typical experiment, several samples of purified host 1 (130 mg each) were separately equilibrated with different amounts (1-70 µL) of guests for 72 h at 298 K in the sealed 15 mL vials, and then their headspace was analyzed. Using this method, the relative vapor pressure (thermodynamic activity) of guest P/P0, where P and P0 are the partial vapor pressure and saturated vapor pressure of guest, respectively, and the guest uptake A (moles of guest per mole of host) were determined. The error in P/P0 determination varied from 5% (for P/P0 > 0.5) to 10% (for P/P0 < 0.1). Nearly half of this error is systematic and corresponds to the error of headspace analysis for a sample of pure liquid guest. The accuracy of the guest uptake determination was (5%. The sorption isotherms for the most studied host-guest systems were determined twice for the same host samples purified from the bound guest as described above. QCM Sensor Study of Host-Guest Binding. In the present study, 10 MHz QCM oscillators were used from ICM Co., Oklahoma City, OK. They are of thickness shear mode, with polished 0.201 in. gold electrodes over 1.2 in. quartz crystals. The thermostated sensor device has four oscillators mounted on polytetrafluoroethylene cover of cylindrical glass cell, with the inner volume of 53 mL and height of 31 mm (Figure 1). Three oscillators were coated with host 1 layer prepared by dropping of 1 µL of host solution in toluene (1 mg/mL) on the center of gold electrodes. This coating (1 µg), after toluene removal, gives a ∼1200 Hz decrease in the oscillator frequency. The average thickness of the host layer on the gold surface was ∼80 nm. This value was estimated by the layer area, mass, and density of calixarene clathrate with toluene, 1.155 g/cm3, calculated from X-ray monocrystal data.

Yakimova et al.

Figure 1. Scheme of measuring part of QCM microbalance: 1, thermostat; 2, polytetrafluoroethylene cover; 3, dosing hole; 4, QCM oscillators; 5, sensor cell.

To remove toluene from initial coating and to regenerate host 1 layer after guest-binding experiment, the oscillators were dried by the air purge at 45 ( 1 °C. Then, the coated oscillators were equilibrated with the ethanol vapor P/P0 ) 0.75 at 298 K until the constant sensor frequency was achieved and dried once more by the air purge. This procedure was repeated until two successive regeneration cycles gave the same frequency of coated oscillators. The temperature, 45 ( 1 °C, and the time of air purge, 120 s, were the same for each step of sensor regeneration and for every studied guest. This time period was sufficient to return the frequency to the initial value for sensor saturated with ethanol vapor after several regeneration cycles. In a typical QCM sensor experiment, a liquid guest was sampled with microsyringe to the cell bottom through the dosing hole in the cell cover. The sampled guest amount was 50% larger than necessary to create their saturation vapor in the sealed cell. Still, the cell was made not hermetical during sensor experiment to avoid capillary condensation of guest on the host surface. The guest relative vapor pressure P/P0 was kept constant but below saturation level by the dosing hole in the cell cover. This level, in dynamic equilibrium, was equal to P/P0 ) 0.75 ( 0.05. The value of P was determined by sampling 100 µL of sensor headspace to GC using a gastight syringe. The values of P0 at 298 K for studied guests were taken from the literature.17 The vapor of liquid water sampled in excess reduces the sensor frequency on 4 ( 1 Hz for both guest-free host and the initial coating of its clathrate with toluene. The corresponding increase of coating mass, 0.3%, is negligibly small compared with the sensor response to the other studied guests. The stoichiometry of host-guest clathrates was determined using the QCM sensor method with the error of 5%. Simultaneous Thermogravimetry and Differential Scanning Calorimetry (TG-DSC) of Host-Guest Inclusion Compounds. Simultaneous TG-DSC analysis was performed using a thermoanalyzer STA 449 C Jupiter (Netzsch) with the temperature rate 10 K/min in an argon atmosphere with the total flow rate of 20 mL/min. For this, 7 mg samples of host-guest clathrates were prepared through the vapor saturation as described above, in the aluminum crucibles (40 µL) with lids having three holes, each of 0.5 mm in diameter. The samples were analyzed by TG-DSC techniques 15-20 min after the clathrates were removed from the hermetically closed vial with saturated guest vapor. During these 15-20 min, clathrate sample was purged by argon flow of 20 mL/min at room temperature on the sample holder of thermoanalyzer. The error of TG experiment is 2-5% depending on the clathrate stability. Initial host 1 (R-phase) is stable in an argon atmosphere up to 400 °C. The TG-DSC data for the guest-free 1 did not show mass

Adamantylcalix[4]arene in QCM Sensor

J. Phys. Chem. B, Vol. 112, No. 49, 2008 15571

loss and any exothermic or endothermic transitions below this temperature (see ESI for further details). The contents of ethanol in the host layer after the “guest exchange-drying” regeneration cycle was determined in the control experiment. For this, the coating with the average thickness of 1000 nm was prepared by drying the host 1 solution in toluene on the flat glass surface (70 cm2). This host layer was equilibrated with the saturated ethanol vapor for 3 days with intermediate drying by hot air at 45 °C for 10 min twice a day. The regenerated coating was scratched from the glass and studied using an STA 449 C Jupiter thermoanalyzer coupled with a quadrupolar mass spectrometer QMS 403 CF Aeolos in an argon atmosphere with total flow rate 75 mL min-1. The heating rate was 10 K min-1. X-ray Diffraction Experiments. Powder X-ray diffraction (XRPD) data were collected using Bruker D8 Advance diffractometer equipped with Vario attachment and Vantec linear position-sensitive detector. In this experiment, Cu KR radiation (40 kV, 40 mA) and graphite monochromator were used. The data were determined at room temperature in the reflection mode with a flat-plate sample. The sample was lightly grounded and loaded into a standard sample holder, which was kept spinning (15 rpm) throughout the data collection. Patterns were recorded in the 2Θ range between 3° and 40°, in 0.008° steps, with a step time of 2 s. Ten powder patterns were obtained and summed for each sample. X-ray diffraction experiment for colorless single crystal of 1 with toluene was performed at room temperature with an Enraf-Nonius CAD-4 diffractometer. Crystal data: C82H96O4, M ) 1145, triclinic, a ) 14.609(9), b ) 15.149(4), c ) 17.134(4) Å, V ) 3295.35 Å3, T ) 295 K, space group P-1, Z ) 2, graphite-monochromated Cu KR radiation, ω-scan mode, 2θ < 69.94°, 12 483 reflections, of which 8271 > 2σ(I). The structure was solved by the direct method and refined by full-matrix least squares (SHELX97 software18) against F2 of all data (the number of adjusted parameters is 770). All non-hydrogens atoms were refined with anisotropic displacement parameters. The H atoms were placed in the calculated positions and allowed to ride on their parent atoms (C-H ) 0.93-0.98 Å, O-H ) 0.82 Å; Uiso(H) ) xUeq (parent atom), where x ) 1.5 for attached O and 1.2 for C). As a result of the partly disordered toluene and adamantyl fragments, the R-factor of the structure has rather high value, R(F2 > 2σ(F2)) ) 0.92. CCDC reference No.: 708006. Results and Discussion Vapor Sorption Isotherms. To find the range of guest relative vapor pressures P/P0 for the saturated guest-host clathrates, the vapor sorption isotherms of acetonitrile, propionitrile, n-butyronitrile, chloroform, n-heptane, ethylbenzene, and n-octane vapors by thermally stable guest-free host 1 (Rphase) were determined (Figure 2, ESI). Vapor sorption isotherms of benzene, toluene, cyclohexane, and tetrachloromethane were determined earlier.19 The R-phase of 1 does not adsorb methanol, ethanol, and n-pronanol vapors up to the guest activity P/P0 ) 0.85. All obtained isotherms have a sigmoidal shape with a threshold on guest thermodynamic activity (P/P0) and saturation level of the guest uptake A (mol of guest per 1 mol of host) above this threshold. This isotherm shape, in terms of the Gibbs’ phase rule, corresponds to a phase transition of clathrate formation.2,20,21 Obtained isotherms were fitted to the equation22

A ) SC(P/P0)N/(1 + C(P/P0)N)

(1)

Figure 2. Vapor sorption isotherms of organic compounds by thermally equilibrated powder of adamantylcalix[4]arene (1) at 298 K. Solid lines are fitting curves calculated by eq 1.

where S is the inclusion stoichiometry (guest/host molar ratio), C is the sorption constant, and N is the cooperativity parameter. The points above P/P0 ) 0.9 were not included because above this activity level the capillary condensation of the guest may become significant. The sorption isotherm for ethylbenzene has two steps of clathrate formation; therefore, a sum of two right parts of eq 1 was used for the data fitting. The fitting parameters of sorption isotherms S and N, guest threshold activity at 50% host saturation a0.50S ) exp(-(ln C)/ N), and calculated guest inclusion Gibbs energy ∆Gc are given in Table 1. The value ∆Gc corresponds to Gibbs energy of guest transfer from a standard state of pure liquid to the saturated clathrate:1

∆Gc ) RT

∫01 ln(P/P0) dY ) RT ln a0.5S

(2)

where Y ) A/S is the host saturation extent. For the two-step sorption isotherm of ethylbenzene, the total inclusion Gibbs energy ∆Gc is the weighted average of ∆G(i) c values for separate inclusion steps:

∆Gc )

∑ Si∆G(i)c ∑ Si

(3)

15572 J. Phys. Chem. B, Vol. 112, No. 49, 2008

Yakimova et al.

TABLE 1: Thermodynamic Parameters of Vapor Sorption Isotherms by Thermally Stable r-Phase of Adamantylcalix[4]arene (1) at T ) 298 K and Data of TG-DSC Analysis for Clathrates of Adamantylcalix[4]arene (1) Prepared Using Saturation of Host Powder with Guest Vaporsa guest

MRD (cm3 mol-1)

CH3CN C2H5CN n-C3H7CN CHCl3 C6H6c CCl4 c c-C6H12 c n-C6H14 d C6H5CH3 c n-C7H16 C6H5C2H5 n-C8H18 n-C9H20 d

11.1 16.0 20.4 21.3 26.2 26.4 27.7 29.9 31.1 34.5 35.8 39.2 43.8

N

a0.5S 0.30 0.52 0.63 0.37 0.32 0.47 0.53

11 16 12 9 5 8 15

δ 0.12 0.05 0.08 0.05 0.04 0.02 0.03

0.26 6 0.04 0.40 3 0.04 0.29; 0.52e 30; 15e 0.01 0.45 9 0.02

S 1.87 3.0 2.6 3.1 2.0 4.4 4.2

∆Gc (kJ mol-1) -3.0 -1.6 -1.1 -2.4 -2.8 -1.9 -1.6

Te (°C)

102 139 111; 147; 218 100; 120 164 134 132 124 2.2 -3.4 144 1.72 -2.3 139 1.58 (0.63)f -2.2 (-3.0; -1.6)g 110 2.3 -2.0 96; 155 200

Tcol(°C) 232 227 248 223 222 228 226

∆m (%)

STG

6.89 12.14 15.55 (14.84)h 27.05 13.36 38.26 25.7 14.18 15.94 16.75 16.88 20.88 (13.12)h 10.97

1.73 2.4 2.6 (0.12)i 3.0 1.90 3.9 4.0 1.84 1.98 1.93 1.84 2.2 (0.83)i 0.93

∆He (kJ mol-1) ∆Hcolb (kJ mol-1) 45.1 26.1 62.8 31.2 52.5 52.6 46.3 56.4 66.5 64.3 54.0 27.9; 25.7j 48.4

-18.9 -6.2 -16.5 -8.6 -7.4 -14.7 -10.9

a δ is standard deviation calculated from shortest distances between the experimental points of sorption isotherm and fitting curve (1) as defined in ref 21. b Enthalpy of the exothermic host 1 collapse without a mass loss. c The data on vapor sorption isotherms are from ref 19. d Vapor sorption isotherms were not determined. e Parameters of separate inclusion steps for isotherms fitted using a sum of two equations (1); in parentheses, parameters for separate inclusion/decomposition steps are given. f The guest amount (mol per 1 mol of host) added in the first inclusion step. g ∆Gc values for inclusion steps. h Mass loss before the last decomposition step. i Stoichiometry of clathrate formed before the last decomposition step. j Enthalpies of first and second steps of guest elimination from clathrate.

where Si is the guest mole number included in the ith step of clathrate formation. According to the data derived from sorption isotherms, host 1 clathrates have the higher threshold values of guest inclusion a0.5S and the higher (less negative) inclusion Gibbs energies ∆Gc (Table 1) than tert-butylcalix[4]arene1,15,23 for all studied guests, except for chloroform. TG-DSC Data for Saturated Clathrates. For clathrates prepared by saturation of host 1 (R-phase) powder with guest vapors at P/P0 ) 1, the composition and parameters of thermal stability were determined using simultaneous TG-DSC analysis. Most of the clathrates studied have perform one-step decomposition according to the obtained TG-DSC curves (Figure 3 and ESI). The exceptions are the clathrates of octane and butyronitrile (ESI), which have two clear-cut decomposition steps. For these guests and for ethylbenzene the number of clathrate decomposition steps does not coincide with the number of clathrate formation steps on sorption isotherms (Figure 1). Such behavior of clathrates is quite usual for calixarenes.2,3,5 The TG-DSC experiment gives a mass loss ∆m, inclusion stoichiometry STG, temperature of DTG peak Te, and enthalpies of guest elimination from clathrates ∆He per 1 mol of guest for each decomposition step, temperature of host collapse Tcol, and the enthalpy of this process ∆Hcol per 1 mol of host (Table 1). The stoichiometry values STG obtained are in agreement within the experimental errors with S values calculated from vapor sorption isotherms (Table 1). Although the binding of methanol and ethanol vapors by 1 was detected in this experiment, no stoichiometry was determined for the formed clathrates due to their very low stability. The enthalpy of endothermic guest elimination ∆He from host 1 clathrates is in the same range of 45-65 kJ mol-1, in the most cases, as for clathrates of tert-butylcalix[5]arene2 and tertbutylcalix[6]arene.9 Most of clathrates of host 1 have lower DTG peaks Te corresponding to the last step of guest elimination (Table 1) than the corresponding clathrates of tert-butylcalix[4]arene.2 Therefore, clathrates of 1 are generally less stable, and 1 is more prospective for use in sensors than tertbutylcalix[4]arene. Clathrates of propionitrile, n-butyronitrile, chloroform, tetrachloromethane, cyclohexane, n-hexane, and n-heptane have also an exothermic transition without mass loss (Figure 3b, Table

Figure 3. Data of simultaneous TG/DSC experiment for saturated clathrates of adamantylcalix[4]arene (1) with toluene (a) and tetrachloromethane (b). Heating rate is 10 K min-1.

1, ESI), which corresponds to some collapse of the host phase.9 The clathrates of the studied aromatic hydrocarbons, n-octane, and n-nonane do not exhibit such an effect (Figure 3a, Table 1, ESI). The enthalpy of the observed exothermic collapse, ∆Hcol (Table 1), does not reach such high absolute values as for tertbutylcalix[6]arene.9 The most negative value was observed for n-hexane clathrate with 1, ∆Hcol ) -15 kJ mol-1, whereas for tert-butylcalix[6]arene clathrates this value reaches -37 kJ mol-1.9 If the energies of molecular interactions in these two hosts are comparable, the sheer volume of the collapsed free space in host 1 may be 2.5 times lower than that in tertbutylcalix[6]arene. The existence of host collapse effect is important for sensor applications because a metastable host with uncollapsed free space may have a different shape of guest sorption isotherm than thermally stable host forms.9

Adamantylcalix[4]arene in QCM Sensor

J. Phys. Chem. B, Vol. 112, No. 49, 2008 15573

Figure 4. Molecular packing of 1 · 2C6H5CH3 clathrate: (a) the view along a axis; (b) single cell view along c axis.

In order to understand the effects of crystal packing on the clathrate properties, the X-ray crystal structure of host 1-toluene clathrate was determined. This clathrate has 1 · 2C6H5CH3 composition, which corresponds to the inclusion stoichiometry values STG and S, derived from TG and vapor sorption data, respectively (Table 1). In this clathrate, host 1 has a brick-type head-to-head packing (Figure 4) which somewhat resembles the packing of tertbutylcalix[4]arene clathrate with the same guest.3,24 But, like in monocrystals of loose tert-butylcalix[4]arene after its sublimation,3 molecules of 1 are coupled forming pairs of vandervaals capsules for toluene. In these pairs, host bowls cap each other with an adamantyl group (Figure 4). Such capping is possible because the calixarene bowls are pinched, having distances 8.96 and 11.51 Å between the quaternary C atoms of distal adamantyl groups. Still, C atoms of methylene bridges in macrocycle form almost a perfect square with diagonals of 7.16 and 7.21 Å. Toluene encapsulated inside these host pairs has a large enough opening to move to the position of the second toluene molecule, which is bound interstitially (Figure 4b). Hence, both guests may have more freedom to leave host or to be exchanged for another guest than toluene encapsulated by tert-butylcalix[4]arene bowls, which are capped more tightly by two tert-butyl groups of neighboring host molecules.3 These structural features of 1 · 2C6H5CH3 clathrate explain partially its one-step formation (Figure 2) and decomposition (Figure 3a). Still, no simple relationship can be expected between the clathrate structure and the numbers of its formation and decomposition steps because these numbers are generally not equal for guest binding and elimination (Table 1).2,3,5 Besides, an intermediate clathrate of calixarene may have rather different structure from that of a saturated clathrate.3 Thermally stable guest-free R-phase and β-phase prepared by saturation of R-phase with toluene vapor at 298 K were characterized using XPRD method (ESI). According to the obtained powder diffraction data, these two forms exhibit different crystal packing. XPRD diffractogram of β-phase matches up with that calculated from X-ray monocrystal data for 1 · 2C6H5CH3 clathrate. Hence, the equilibration of host 1 powder with toluene vapor gives the same clathrate as the crystallization from toluene solution. QCM Sensor Data. The QCM sensor experiment was performed according to the scheme given in Figure 5. In this experiment, the sensor responses of guest-free host 1 layer (80 nm) were determined for vapors of various guests with P/P0 ) 0.75 at 298 K (step 1). The sensing of guests, for which the

Figure 5. Scheme of QCM experiment including a response of regenerated sensor to guest vapor with relative vapor pressure of P/P0 ) 0.75 at 298 K (step 1), air purge of saturated clathrate giving its partial decomposition (step 2), response of sensor with partially regenerated host 1 layer to guest vapor (P/P0 ) 0.75) at 298 K (step 2a), guest substitution with ethanol vapor (step 3), and air purge of unstable 1 · yEtOH clathrate (step 4).

enclathration parameters were determined by TG-DSC and headspace GC methods (see above), was monitored. The examples of the observed responses for several guests are shown in Figure 6a. The relative vapor pressure of guests, P/P0 ) 0.75, used in QCM experiment, corresponds to the saturation part of vapor sorption isotherms for 1 powder in all cases, where the host-guest binding was observed below P/P0 ) 0.85 (Figure 2, ESI). The values of sensor frequency changes ∆F1 upon complete saturation of initially guest-free host 1 with guest vapors are given in Table 2. These values were normalized to the same mass of guest-free host 1 layer giving frequency change ∆Fhost ) 1200 Hz. The guest/host molar ratio Sq in the host 1 layer after the guest vapor binding was calculated using the equation

Sq ) (∆F1 /∆Fhost)(Mhost /Mguest) where Mhost and Mguest are molar weights of host and guest, respectively. The calculated Sq values are given in Table 2. These Sq values are in agreement within experimental accuracy with the stoichiometry from sorption isotherms, S, and from TG experiment, STG (Tables 1 and 2). This proves the absence of guest in the initial host 1 layer on the sensor surface in step 1, Figure 5. Such proof is not trivial for the sensing behavior of host 1 because its purge with hot air did not give a guest-free sensor coating. For the most of studied guests, this regeneration procedure (step 2, Figure 5) did not return the sensor frequency to the initial value of step 1. The next sensor experiment with the vapors of the same guest, step 2a, returned sensor frequency

15574 J. Phys. Chem. B, Vol. 112, No. 49, 2008

Yakimova et al. TABLE 2: QCM Sensor Data for the Binding of Guest Vapors (P/P0 ) 0.75) by Thin Layer of Host 1 at 298 Ka guest

Sq

∆F1 (Hz)

∆F2a/∆F1

S2a

CH3OH C2H5OH CH3CN C2H5CN n-C3H7CN CHCl3 C6H6 CCl4 c-C6H12 n-C6H14 C6H5CH3 n-C7H16 C6H5C2H5 n-C8H18 n-C9H20

1.7 1.5 2.0 2.7 2.5 3.4 2.1 4.5 4.1 1.7 2.0 1.9 1.8 2.3 0.9

66 87 102 183 213 488 204 870 432 183 233 242 236 319 145

0.99 1.02 0.84 0.83 0.74 0.73 0.49 0.72 0.75 0.42 0.23 0.61 0.72 0.80 0.30

1.7 1.5 1.7 2.2 1.8 2.4 1.0 3.2 3.1 0.7 0.5 1.2 1.3 1.8 0.3

Sensor responses were normalized to the value of ∆Fhost ) 1200 Hz. a

Figure 6. Responses of QCM sensor coated with guest-free host 1: (a) to organic guest vapors in step 1 after complete regeneration cycle, steps 2-4; (b) to CCl4 vapor for 1 regenerated using air drying at 45 °C (curve II, step 2a) and guest exchange technique (curves I, III, step 1); (c) to ethanol vapor for initial host (solid line, step 1) and in subsequent experiment after intermediate air drying at 45 °C (dotted line, step 2a). Sensor responses were normalized to the value of ∆Fhost ) 1200 Hz; guest relative vapor pressure is P/P0 ) 0.75, T ) 298 K.

to that of saturated clathrate 1 · Sq within the experimental errors. But the corresponding frequency change, ∆F2a, was in the range from 23 to 84% of the ∆F1 value for the most studied guests (Table 2 and Figure 6b). Complete regeneration in step 2 with ∆F2a ) ∆F1 was observed only for methanol and ethanol (Figure 6c and Table 2). By showing a reversible binding to 1, ethanol and methanol may be good agents for host 1 regeneration through the exchange of bound guest. These compounds do not form stable clathrates with thermally stable R-phase of this calixarene. In QCM sensor, a layer of 1 gave a response ∆F1 to methanol

and ethanol vapors that corresponds to guest/host molar ratios of 1.7 and 1.5, respectively (Table 2). The exchange for ethanol, step 3, was observed explicitly for guests with higher inclusion stoichiometry and molar weight. Initially, ethanol increased the weight of the quartz sensor, coated with the air-purged toluene clathrate of 1, probably by filling the free binding sites in host 1. Then the sensor frequency gradually changed to the value corresponding to the much lower mass of ethanol clathrate, ESI. The success of host regeneration using this exchange technique depends on the host layer thickness. It took less than 1 h for the complete elimination of toluene from the 1 · 2C6H5CH3 clathrate with the layer thickness of 80 nm on QCM sensor. But even 3 days was not enough to exchange and remove guest from the layer of the same clathrate having the average thickness of 1000 nm. In this independent control experiment with the clathrate layer on the flat glass surface, only 60% of bound toluene was removed under comparable conditions, according to the results of TG-MS analysis, ESI. The determined residual contents of ethanol in the sample was below 2 mol %. It is evident that a thicker layer of host makes the rate of guest diffusion and exchange much slower. Therefore, substantially lower contents of ethanol after regeneration procedure may be expected for the 80 nm thick layer of host 1. This small amount of ethanol should not produce significant effects on the subsequent binding properties of the host in sensor. Once preparation of guest-free host layer on the sensor surface is made possible, the frequency changes in steps 2a and 1 can be determined, and the ∆F2a/∆F1 parameter can be used as a measure of the guest binding reversibility. This value is different for different guests, reflecting the difference in the guest structures and guest-host complementarity in clathrates. As such, the ∆F2a/∆F1 value may be used as a relative measure for recognition of guest vapors. Remarkably, this parameter allows distinguishing between n-heptane and toluene vapors, which produced nearly the same response of the sensor with guest-free host 1, ∆F1, but showed completely different reversibility parameters 0.61 and 0.23, respectively (Table 2). The same distinctions were observed for propionitrile and n-hexane, toluene and ethylbenzene, and n-butyronitrile and benzene pairs. For each pair, the difference of ∆F2a/∆F1 values exceeds experimental errors, whereas guestfree host 1 gives the same sensor response, ∆F1. Overall, 7 and 6 out of 15 studied guests were successfully distinguished using separately the sensor responses of com-

Adamantylcalix[4]arene in QCM Sensor

J. Phys. Chem. B, Vol. 112, No. 49, 2008 15575 edge Dr. Aidar Gubaidullin, A.E. Arbuzov Institute of Organic and Physical Chemistry, Kazan, Russia, for XRPD experiments. Supporting Information Available: Powder X-ray diffraction data, data of simultaneous thermogravimetry and differential scanning calorimetry for adamantylcalix[4]arene and its clathrates, QCM data for guest exchange in toluene clathrate of adamantylcalix[4]arene, data of thermogravimetry with mass spectrometric analysis of released vapors for product of guest exchange in this clathrate, and vapor sorption isotherms of guests by R-phase of adamantylcalix[4]arene. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes

Figure 7. Molecular recognition of guest vapors using QCM-sensor responses of completely regenerated host 1, ∆F1 (step 1), and of partially regenerated host, ∆F2a (step 2a). Dotted line corresponds to the points with reversible guest-host binding. Deviation from this line corresponds to the value of binding reversibility parameter, R ) ln(∆F2a/ ∆F1). Point size corresponds to the value of experimental errors.

pletely regenerated host, ∆F1, and those of partially regenerated host, ∆F2a, respectively. But the combination of these two parameters allows an accurate recognition of 13 guests (Figure 7). Only the responses of n-butyronitrile and ethylbenzene are overlapped within the experimental accuracy ((5% for each point). Considering that both ∆F1 and ∆F2a/∆F1 parameters can be obtained from a single host 1-based QCM sensor in a single set of experiments, this single sensor device has the recognition capability of two-sensor array. Conclusions The present study gives a guideline for application of calixarene or any other clathrate-forming host in vapor sensors. Guest binding cooperativity reduces reversibility of this process, sometimes rendering complete regeneration of sensor by simple purging with hot air practically impossible. Still a thin layer of calixarene in sensors may be regenerated using a special procedure with guest exchange. This makes guest binding in sensors virtually reversible, removing possible memory effects caused by the residuals of the previously bound guests. The sensor response of the completely regenerated host, once quantitatively estimated, allows calculating the value of binding reversibility parameter for the experiment, where host purge by air gives only its partial regeneration. This parameter gives an additional dimension to the quantitative picture of molecular recognition for a single sensor device. Hence, a kind of “breathin-breath-out” recognition technique may be used in detection of organic vapors. Acknowledgment. This research was supported by the RFBR (No. 08-03-01107-a`) and BRHE (REC007). Authors acknowl-

(1) Gorbatchuk, V. V.; Tsifarkin, A. G.; Antipin, I. S.; Solomonov, B. N.; Konovalov, A. I.; Lhotak, P.; Stibor, I. J. Phys. Chem. B 2002, 106, 5845–5851. (2) Ziganshin, M. A.; Yakimov, A. V.; Safina, G. D.; Solovieva, S. E.; Antipin, I. S.; Gorbatchuk, V. V. Org. Biomol. Chem. 2007, 5, 1472–1478. (3) Atwood, J. L.; Barbour, L. J.; Jerga, A. Chem. Commun. 2002, 2952–2953. (4) Benevelli, F.; Kolodziejski, W.; Wozniak, K.; Klinowski, J. Phys. Chem. Chem. Phys. 2001, 3, 1762–1768. (5) Brouwer, E. B.; Enright, G. D.; Udachin, K. A.; Lang, S.; Ooms, K. J.; Halchuk, P. A.; Ripmeester, J. A. Chem. Commun. 2003, 1416–1417. (6) Kalchenko, V. I.; Koshets, I. A.; Matsas, E. P.; Kopylov, O. N.; Solovyov, A.; Kazantseva, Z. I.; Shirshov, Yu. M. Mater. Sci. 2002, 20, 73–88. (7) Dickert, F. L.; Schuster, O. AdV. Mater. 1993, 5, 826–829. (8) Ripmeester, J. A.; Enright, G. D.; Ratcliffe, C. I.; Udachin, K. A.; Moudrakovski, I. L. Chem. Commun. 2006, 4986–4996. (9) Yakimov, A. V.; Ziganshin, M. A.; Gubaidullin, A. T.; Gorbatchuk, V. V. Org. Biomol. Chem. 2008, 6, 982–985. (10) Schatz, J.; Schildbach, F.; Lentz, A.; Rastatter, S. J. Chem. Soc., Perkin Trans. 2 1998, 75–77. (11) Pirondini, L.; Dalcanale, E. Chem. Soc. ReV. 2007, 36, 695–706. (12) Ananchenko, G. S.; Udachin, K. A.; Dubes, A.; Ripmeester, J. A.; Perrier, T.; Coleman, A. W. Angew. Chem. 2006, 118, 1615–1618. (13) Ziganshin, M. A.; Yakimova, L. S.; Khayarov, K. R.; Gorbatchuk, V. V.; Vysotsky, M. O.; Bo¨hmer, V. Chem. Commun. 2006, 3897–3899. (14) Khomich, N.; Shokova, E. A.; Kovalev, V. V. Synlett 1994, 1027– 1028. (15) Gorbatchuk, V. V.; Tsifarkin, A. G.; Antipin, I. S.; Solomonov, B. N.; Konovalov, A. I.; Seidel, J.; Baitalov, F. J. Chem. Soc., Perkin Trans. 2 2000, 2287–2294. (16) Ziganshin, M. A.; Yakimov, A. V.; Antipin, I. S.; Konovalov, A. I.; Gorbatchuk, V. V. Russ. Chem. Bull. 2004, 53, 1478–1543. (17) Boublik, T.; Hala, V.; Fried, E. The Vapour Pressures of Pure Substances; Elsevier: Amsterdam, 1973. (18) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112–122. (19) Gorbatchuk, V. V.; Savelyeva, L. S.; Ziganshin, M. A.; Antipin, I. S.; Sidorov, V. A. Russ. Chem. Bull. 2004, 53, 60–65. (20) Dewa, T.; Endo, K.; Aoyama, Y. J. Am. Chem. Soc. 1998, 120, 8933–8940. (21) Gorbatchuk, V. V.; Tsifarkin, A. G.; Antipin, I. S.; Solomonov, B. N.; Konovalov, A. I. MendeleeV Commun. 1999, 11–13. (22) Edsall, J. T.; Gutfreund, H. Biothermodynamics; Wiley: New York, 1983. (23) Gorbatchuk, V. V.; Tsifarkin, A. G.; Antipin, I. S.; Solomonov, B. N.; Konovalov, A. I. J. Inclusion Phenom. Macrocyclic Chem. 1999, 35, 389–396. (24) Andreetti, G. D.; Ungaro, R.; Pochini, A. J. Chem. Soc., Chem. Commun. 1979, 1005–1007.

JP804277U