Highly sensitive nitric oxide detection using X-ray photoelectron spectroscopy.

Nitric oxide (NO) is important physiologically as a gaseous messenger in the central nervous system;1 its level is also associated with several disease states.2 The concentration of NO in these cases is in the nanomolar range; therefore accurate and direct detection of NO at these levels is important. NO is also a decomposition product of several explosives3 so detection of trace amounts of explosive, for example in security screening, also requires a sensitive NO assay. Thus it is not surprising that a large variety of NO detection systems have been discussed for implementation in these applications. NO is now detected using fluorescence,4 chemiluminescence,5 electrochemistry,6 electron paramagnetic resonance spectrometry,7 and electrical8 methods. These techniques employ surface-modified metals or semiconductors, usually as electrodes, that are treated chemically or physically to increase their interaction with NO. A range of strategies has been used for modifying these electrodes for this purpose. One common surface modification technique is to coat the electrode with a metalaporphyrin: NO binds to the iron atom of the hemelike molecule;9 experimental and computational studies have elucidated the geometric and vibrational features of heme-NO binding.10 Iron porphyrins are more reactive to NO than either to O2 or CO,11 which makes them a good choice for use in detection of NO in a competitive environment where CO or O2 could be present. Clearly, the stability of the interface between a sensor device and the NO-binding surface heme complex is critical for device lifetime, as is the quantitative control of surface modification for measurement reproducibility. We now report a new method to stably bond monolayers of an iron heme complex onto silicon that enables direct, quantitative detection of NO at low levels. We have described bonding self-assembled monolayers of organophosphates onto oxide surfaces;12 these monolayer interfaces are stable to air and water.13 In particular we have found these monolayers to be dense and ordered films on oxide terminated silicon.14 We used a monolayer formed from 11-hydroxyundecylphosphonic acid (1) on SiO2/Si as our reactive platform to attach NO-bonding molecule hematin. Several drops of aq HCl were added to a 10 μM solution of hematin (Aldrich) in CH2Cl2, which was then heated with a 60 μM solution of SOCl2 under inert atmosphere for 24 h to give its diacyl chloride, 3. That both carboxylic acid groups of the hematin were converted to the acyl chloride was confirmed by IR, which showed replacement of ν(CdO)-OH (1710 cm-1) by ν(CdO)-Cl (1800 cm-1) (see Supporting Information). Heme-terminated surface 4 was then obtained by treating 2 with a solution of 3 in dry CH2Cl2 under inert atmosphere for 24 h, followed by sonication successively with CH2Cl2 and water. Coupons of 4 were then transferred into ultrahigh vacuum (UHV) for X-ray photoelectron spectroscopic analysis (XPS). The survey scan of 4 showed distinct peaks in the N1s and Fe2p regions, which confirms the attachment of hematin on the surface (see Supporting Information). Surface 4 was then treated with NO to give adduct 5 (Scheme 1). Caution: NO is toxic, so care must be taken to work in a well-ventilated hood. The quantitative interaction of NO with 4 was studied by XPS. Pure nitric oxide (MGI) was passed over 4 in a chamber that had been purged with argon to remove oxygen to avoid formation of NO2. The resulting adduct 5 was then transferred into UHV (base pressure, 5 × 10-9 Torr) for XPS analysis. The survey spectrum of 5 was similar to that of 4 except that two new peaks appeared in the N1s region. In contrast to 4, which shows a single peak, N2, (BE ) 399.5 eV) in the N1s region attributed to the nitrogens of the hematin molecule, the N1s region of 5 has three peaks (Figure 1). The peaks at 398 (N1) and 399.5 eV are attributed to nitrogens from 5 and from unreacted 4, respectively, and the peak at 405.2 eV (N3) is attributed to the N1s of iron bound NO. The shift to lower binding energy for the heme nitrogens of 5 as compared to 4 can be explained by noting that NO, a free radical, bonds to iron as the nitrosyl ligand by a formal 1 ereduction of the metal.15 A similar decrease in the binding energy of the Fe2p peak was also observed (Figure 2). The ratio of peak areas for N3 and N1 (Figure 1) is 1:4, consistent with the expected stoichiometry of the hematin-NO complex 5. The sum of areas under the N1 and N2 peaks of 5 equals the area of the peak under the N1s peak of 4, which confirms that the peak at 405.2 eV is in fact due to NO attachment and not to the degradation of the hematin complex (see Supporting Information). Quartz crystal microgravimmetry (QCM) studies16 showed that the coverage of 1 on SiO2/Si is 850 pmol/cm2 ((5%), and the loading of 3 on 1 is 300 pmol/cm2 ((6%). Before we could estimate Scheme 1. Schematic of the Surface Modification Strategy Published on Web 05/11/2007