Using Hybrid Organic − Inorganic Surface Technology to Mitigate Analyte Interactions with Metal Surfaces in UHPLC

: Interactions of analytes with metal surfaces in high-performance liquid chromatography (HPLC) instruments and columns have been reported to cause deleterious e ﬀ ects ranging from peak tailing to a complete loss of the analyte signal. These e ﬀ ects are due to the adsorption of certain analytes on the metal oxide layer on the surface of the metal components. We have developed a novel surface modi ﬁ cation technology and applied it to the metal components in ultra-HPLC (UHPLC) instruments and columns to mitigate these interactions. A hybrid organic − inorganic surface, based on an ethylene-bridged siloxane chemistry, was developed for use with reversed-phase and hydrophilic interaction chromatography. We have characterized the performance of UHPLC instruments and columns that incorporate this surface technology and compared the results with those obtained using their conventional counterparts. We demonstrate improved performance when using the hybrid surface technology for separations of nucleotides, a phosphopeptide, and an oligonucleotide. The hybrid surface technology was found to result in higher and more consistent analyte peak areas and improved peak shape, particularly when using low analyte mass loads and acidic mobile phases. Reduced abundances of iron adducts in the mass spectrum of a peptide were also observed when using UHPLC systems and columns that incorporate hybrid surface technology. These results suggest that this technology will be particularly bene ﬁ cial in UHPLC/mass spectrometry investigations of metal-sensitive analytes.

F or the last fifty years, stainless steel has been the most commonly used construction material for high-performance liquid chromatography (HPLC) instruments and columns. The combination of high strength, compatibility with a wide range of chemicals, manufacturability, and low cost makes it an excellent material for many applications. However, stainless steel can negatively impact the peak shape and recovery of certain compounds. 1−11 Analytes that show these effects typically contain acidic functional groups, such as phosphate and carboxylate moieties, and the severity increases with the number of these groups in the molecule. Stainless steel is susceptible to corrosion, particularly when exposed to acidic and/or halide-containing mobile phases, 12 and corroded surfaces may be particularly prone to interact with certain analytes. 6 Stainless steel may also release iron ions into the mobile phase, and these ions can bind to the stationary phase and interact with metal-sensitive analytes. 3 These interactions include complexation, 1 oxidation, 13 and epimerization reactions. 3 Alternative metals such as titanium and MP35N (a nickel− cobalt alloy) have been used for some applications because of their improved corrosion resistance, 14 but still cause deleterious chromatographic effects under certain conditions. 15 The engineering plastic polyether ether ketone (PEEK) has been employed to avoid these effects, but suffers from limited pressure resistance and some solvent incompatibilities. 16 PEEK is also relatively hydrophobic and may require conditioning to avoid losses of hydrophobic analytes. 17,18 An alternative approach to mitigate interactions of analytes with metal surfaces is to add chelators such as ethylenediaminetetraacetic acid to the mobile phase or sample. 1,5,7,10,19 Other chelators such as citric acid, acetylacetone, and medronic acid have been used for LC/MS analyses. 20−22 However, the use of chelators can cause ion suppression as well as problems with subsequent analyses due to the persistence of the additive in the HPLC system. 23 To address these issues, we have explored the use of a hybrid organic−inorganic surface modification, applied using vapor deposition to the metal components in ultra-HPLC (UHPLC) instruments and columns. A hybrid surface based on an ethylene-bridged siloxane polymer has been found to be well-suited for reversedphase and hydrophilic interaction chromatography. 24 This differs from the silica and carbosilane/carboxysilane vapor deposition coatings that have been reported by others. 25 Here, we describe evaluations of the performance of UHPLC instruments and columns incorporating this hybrid surface technology (HST) relative to their conventional counterparts.
Instrumentation. A rame-hart (Succasunna, NJ) model 190 CA Goniometer was used to measure contact angles on silicon wafers to which the hybrid surface was applied. All UHPLC and MS instrumentation, columns, and data systems were from Waters Corp. (Milford, MA). UHPLC columns and instrumentation incorporating HST are commercially available and carry the PREMIER brand name.
Frit Testing. The frit tests were performed using an ACQUITY UPLC I-Class system consisting of a binary solvent manager, fixed loop sample manager, CH-A column heater, and an ACQUITY tunable UV (TUV) detector. To eliminate the system as a variable in the testing, a PEEK needle, a PEEK sample loop, and an active preheater with the hybrid surface were used. A custom fixture was used to allow testing of individual frits without a column. The mobile phase was aqueous 10 mM NH 4 OAc (pH 6.8), chosen to avoid hydrolysis of ATP. A flow rate of 0.2 mL/min and a temperature of 30°C were used. ATP was monitored using absorbance at a wavelength of 260 nm.
Chromatographic ConditionsNucleotide Separations with UV Detection. We used ACQUITY UPLC H-Class instruments equipped with a quaternary solvent manager, a flow-through-needle sample manager (SM-FTN), a CH-A column heater, and an ACQUITY photodiode array detector. Both standard instruments and modified versions using components treated with the HST were used. Isocratic separations of ATP and AMP were achieved using aqueous 10 mM NH 4 OAc mobile phases, at a flow rate of 0.5 mL/min. Unless noted otherwise, the pH of the mobile phase was 6.8. The samples, freshly prepared daily in 100% water, were injected into ACQUITY UPLC BEH C 18 1.7 μm, 2.1 × 50 mm columns at 30°C. Separations were also performed using columns of the same dimensions constructed with hardware modified with the HST and packed with the same batch of the stationary phase. The injected masses ranged from 20 to 100 ng of each nucleotide. All tests were carried out using new columns. Columns were equilibrated with the mobile phase before the injections. The UV response at 260 nm was recorded using an Empower 3 Chromatography Data System.
Chromatographic ConditionsNucleotide Separations with Tandem MS Detection. We used a HSTmodified ACQUITY UPLC I-Class instrument equipped with a binary solvent manager, a SM-FTN, a CH-A column heater, and a standard Xevo TQ-XS mass spectrometer. Gradient separations were carried out using aqueous 10 mM NH 4 OAc (pH 6.8) and ACN. The gradient conditions are given in the Supporting Information. The flow rate was 0.5 mL/min, and the columns were maintained at 35°C. Separations were performed using a standard ACQUITY UPLC HSS T3 1.8 μm, 2.1 × 50 mm column and a column constructed with hardware modified with the HST and packed with the same stationary phase. Samples containing ATP, ADP, AMP, and adenosine were injected at mass loads ranging from 20 fg to 5 ng. Negative-ion electrospray mode was used with a capillary voltage of −0.5 kV, a desolvation temperature of 600°C, a desolvation gas flow of 1000 L/h, a cone gas flow of 150 L/h, and a nebulizer gas pressure of 7.0 bar. The SRM conditions are given in the Supporting Information.
Chromatographic ConditionsPeptide Separations. An equimolar mixture of angiotensin I, enolase T37, and doubly phosphorylated IR peptide were analyzed by LC-UV using a HST-modified ACQUITY UPLC H-Class Bio system equipped with an ACQUITY UPLC TUV detector. Separations were performed on standard ACQUITY UPLC CSH C 18 1.7 μm, 2.1 × 50 mm columns using 0.1% FA in water (mobile phase A) and 0.09% FA in ACN (mobile phase B). Separations were also performed using columns of the same dimensions constructed with hardware modified with the HST and packed with the same batch of the stationary phase as the standard columns. Three columns of each type were tested using a temperature of 60°C, a flow rate of 0.2 mL/min, and a gradient from 0.5 to 40% B in 12 min, followed by 40−80% B in 2 min. The columns were tested before and after conditioning with 4 nmol injections of the IR peptide. Prior to conditioning, four injections of the three-peptide mixture at a mass load of 20 pmol were run, and one injection at the same mass load was analyzed after conditioning and a water blank. Analyses were performed with UV detection at 214 nm using MassLynx 4.1 and Empower 3 software for data acquisition. UNIFI 1.8 was used for data analysis. A Xevo G2-XS QTOF was used for MS detection, using a capillary voltage of 2.5 kV, a sampling cone and source offset of 80, a source temperature of 120°C, a desolvation temperature of 500°C, a desolvation gas flow of 800 L/h, and a collision energy of 10 eV. The instrument was operated in positive-ion electrospray mode.
Chromatographic ConditionsOligonucleotide Separations. Trecovirsen was analyzed by LC-UV using the same HST-modified ACQUITY UPLC H-Class Bio system used for the peptide separations. Separations were performed on standard ACQUITY UPLC Oligonucleotide BEH C 18 1.7 μm 2.1 × 50 mm columns, using a mobile phase containing 15 mM triethylamine and 400 mM HFIP in water (mobile phase A) and a 50:50 (v/v) solution of mobile phase A and methanol (mobile phase B). Separations were also performed using columns of the same dimensions constructed with hardware modified with the HST and packed with the same batch of the stationary phase. Three columns of each type were tested. Separations were run at a temperature of 60°C, a flow rate of 0.2 mL/min, and a gradient from 0.5 to 40% B in 12 min, followed by 40−80% B in 2 min. Four injections of trecovirsen were made at a mass load of 1.5 ng. Next, the columns were conditioned by injecting 0.2 μg of trecovirsen followed by 10 μL of water to ensure no carryover of the oligonucleotide. This was followed by a fifth 1.5 ng injection of trecovirsen to assess the separation on the conditioned columns. Analyses were performed with UV detection at 260 nm using MassLynx 4.1 for data acquisition and UNIFI 1.8 for data analysis. A Xevo TQ-XS mass spectrometer was used for MS detection with a capillary voltage of 2.0 kV, sampling cone at 45, source offset at 30, a source temperature of 150°C, a desolvation temperature of 600°C, desolvation gas flow set at 1000 L/h, and a collision energy of 5 eV.

■ RESULTS AND DISCUSSION
Characterization of the Hybrid Surface. The surface we developed is composed of an ethylene-bridged siloxane polymer [(O 1.5 SiCH 2 CH 2 SiO 1.5 ) n ] that is formed on metal substrates using a vapor deposition process. 26 The chemical composition of this surface is related to that of ethylenebridged hybrid (BEH) chromatographic particles. 27 We have measured a static water contact angle of approximately 30°for the ethylene-bridged siloxane surface on a silicon wafer, which is significantly lower than the 70−90°reported for PEEK. 28 This indicates that the hybrid surface is more hydrophilic than PEEK, making it less prone to hydrophobic adsorption. The vapor deposition technique is able to provide an effective surface even on high aspect ratio substrates, such as tubing with an internal diameter of 100 μm and a length of 368 mm (see Supporting Information). This makes it possible to implement the technology across diverse types of LC instruments and column components.
To characterize the effectiveness of this surface in mitigating interactions of metal-sensitive analytes with metal substrates, we measured the recovery of a low mass load of ATP, which is known to show severe losses when using metal components. 4,6,10,11 Column frits with and without the hybrid surface were in turn placed in a holder and connected to a UHPLC system with PEEK tubing used in place of stainless steel tubing. No column was used in this experiment. The mobile phase was aqueous 10 mM NH 4 OAc (pH 6.8). This pH was chosen to avoid hydrolysis of ATP. A UV detector was used, and the area of the peak was compared to that obtained without the frit, allowing the recovery to be calculated. Titanium frits with a diameter of 4.6 mm, a thickness of 1.5 mm, and a porosity grade of 0.2 μm gave an ATP recovery of less than 5% for a 10 ng injection. After the hybrid surface was applied, the ATP recovery increased to an average of 99.7% with a standard deviation of 1.6% for 32 frits, each prepared with an independent application of the hybrid surface. This demonstrates the effectiveness and reproducibility of the HST. We note that column frits account for a large proportion of the metal surface area accessible to the analyte in typical UHPLC systems (ca. 52%). However, to minimize adsorption, all metal surfaces that the analyte contacts need to be treated. 26 Characterization of a UHPLC Instrument and Columns Incorporating the Hybrid Surface. We next compared the separations of ATP and AMP achieved using a standard UHPLC instrument and column versus their counterparts constructed using hardware treated with the hybrid surface. A series of fifteen injections were made of a solution containing 20 ng of each nucleotide, using an aqueous 10 mM NH 4 OAc (pH 6.8) mobile phase. Shown in Figure 1 are chromatograms for the fifteenth injections obtained using (i) a standard system and column, (ii) a standard system and a column containing the same packing material but using column hardware treated with the hybrid surface, and (iii) a system and column treated with HST. The results show that AMP is only moderately sensitive to metal surfaces, with modest changes seen across the different system and column combinations. With the standard system and column, the AMP peak was noticeably broader and tailed more than with the other combinations, and the peak area was lower. In contrast, ATP is highly sensitive to metal surfaces. With the standard system and column, ATP was not detected in any of the injections. When an HST column was used with the standard system, an ATP peak was observed, but it tailed and had an area that was much lower than expected. For this configuration, the area and height of the ATP peak increased with repeated injections. When both the HST system and column were used, the ATP peak was much narrower, more symmetric, and consistent from the first to the fifteenth injection. This shows that the best performance for highly metal-sensitive analytes is obtained using HST in both the UHPLC column and the system.
pH Dependence of ATP Losses. We investigated the dependence of the peak area for ATP on mobile phase pH, using both a standard column and a column constructed using hardware treated with the hybrid surface. The UHPLC system used for this experiment had components that were treated with the hybrid surface. All the mobile phases contained aqueous 10 mM NH 4 OAc, with the pH adjusted by adding either acetic acid or ammonium hydroxide. A series of 50 sequential injections of 100 ng of ATP was made on 1.7 μm BEH C 18 2.1 × 50 mm columns. The results are shown in Figure 2. Using a pH 4.5 mobile phase, we observed an almost complete loss of ATP in the initial injections on the standard column. As more injections were made, the peak area gradually increased, but never reached the expected area (the area observed using the HST column), even after 50 injections. Using a pH 6.8 mobile phase, the standard column showed a ca. 50% ATP loss in the first injection, with a gradual increase in peak area with increasing injection number. When a pH 8.5 mobile phase was used with a standard column (data not shown), the ATP peak area was slightly low (ca. 5% loss) in the first injection, but quickly reached the expected area. This trend of reduced losses with increasing mobile phase pH is in agreement with prior studies. 6,8,10 It has been hypothesized that the pH dependence is due to changes in the surface charge of the oxide layer on the metal surface, with an isoelectric point (pI) of approximately 7 for 316 stainless steel. 29 At pH values below this pI, the surface oxide layer is positively charged and may bind analytes that are negatively charged through electrostatic interactions. In contrast, columns constructed with hardware treated with the hybrid surface showed high and Figure 2. Comparison of the peak area of ATP vs injection number using different mobile phase pH values for a standard 1.7 μm BEH C 18 column and a BEH C 18 column constructed with hardware treated with the hybrid surface. The mobile phases contained 10 mM NH 4 OAc, with the pH adjusted to either 4.5 or 6.8. Detection was by absorbance at 260 nm. Fifty sequential injections of 100 ng of ATP were made. The UHPLC system employed for this experiment had a flow path treated with the hybrid surface. much more consistent ATP peak areas, regardless of the mobile phase pH.
Mass Load Dependence of ATP Recovery. We also investigated the dependence of the peak area of ATP on the mass injected for a standard column, using a UHPLC system treated with the hybrid surface. For this experiment, we used an aqueous 10 mM NH 4 OAc (pH 6.8) mobile phase. Fifty sequential injections of ATP were made on a standard 1.7 μm BEH C 18 2.1 × 50 mm column. The experiments were carried out for two different injection masses: 25 and 100 ng. The results presented in Figure 3 show that at the higher loading, the peak area for the first injection was less than half of the value expected based on the peak area observed using a column constructed with the hybrid surface (indicated by the solid purple line). The peak area increased to 80% of the expected value by the seventh injection, slowly continuing to increase to 85% after 50 injections. At the lower loading, ATP was not detected in the first three injections, and the peak area increased very gradually, only reaching 35% of the expected value (indicated by the dashed purple line) after 50 injections. These results show that analyte losses were more severe for the lower on-column mass loading, in agreement with studies reported by others. 11 They also suggest that the interaction of ATP with the metal surfaces is partially reversible, because the peak areas never reached the expected values, even after 50 injections. It appears that some of the adsorbed analyte is released as the mobile phase continually flows through the column. This result indicates that attempting to condition an HPLC system and column by making injections of the analyte before the sample to be analyzed may fail, because the analyte adsorbed in the conditioning injections may be partially eluted before the sample is injected.
To investigate the behavior at lower mass loads, we carried out measurements using LC/MS/MS. Samples containing adenosine, AMP, ADP, and ATP were analyzed at on-column loads ranging from 20 fg to 5 ng of each analyte. A UHPLC system treated with the hybrid surface was used, and the results were compared for 2.1 × 50 mm standard and HST columns packed with 1.8 μm HSS T3. ACN gradient separations were carried out, with an aqueous mobile phase containing 10 mM NH 4 OAc (pH 6.8). Using the standard column, ATP and ADP were not detected over the entire range of loads, while AMP gave a linear response from 5 pg to 2 ng and adenosine from 200 fg to 0.5 ng (see Supporting Information). In contrast, using the column constructed with the hybrid surface, AMP gave a linear response from 100 fg to 2 ng, ADP from 500 fg to 5 ng, ATP from 2 pg to 5 ng, and adenosine from 20 fg to 0.5 ng (see Supporting Information). Shown in Figure 4 are overlaid chromatograms for the four compounds at a mass load of 50 pg on the two columns. For the HST column, all four analytes were detected, with high signal-to-noise ratios. In contrast, for the standard column, ATP and ADP were not observed. While AMP was detected, the peak was considerably broader and less symmetric than that observed using the hybrid surface column. Adenosine gave similar results with the two different columns.
Evaluation for Peptide Separations. Next, we investigated the utility of UHPLC instruments and columns constructed with the hybrid surface for the analysis of peptides. An equimolar mixture of the three peptides (angiotensin I, enolase T37, and a doubly phosphorylated IR peptide) was separated using either standard columns or columns constructed with the hybrid surface (n = 3 of each), all packed with the same batch of stationary phase. A UHPLC system modified with the hybrid surface was used, with both UV and MS detection, employing an ACN gradient with mobile phases containing 0.1% FA. The initial column performance was evaluated from the first four injections, using a mass load of 20 pmol (25−50 ng) of each peptide. Then, a high mass load (4 nmol, 7.1 μg) of the IR peptide was injected to condition the columns, and a fifth injection of the peptide mixture at the 20 pmol load was made to determine the impact of conditioning. Representative UV chromatograms resulting from the fourth and fifth injections are shown in Figure 5. While the peak areas for angiotensin I and enolase T37 were found to be similar across the first four injections for both column types, the doubly phosphorylated IR peptide gave extremely low peak areas with the standard columns ( Figure 5A). In comparison, the hybrid surface columns showed reproducible performance over the five injections regardless of column conditioning ( Figure 5B). The post-conditioning changes in the peak area were less than 3% with the hybrid surface columns. However, with the standard columns, it was only after conditioning that the IR peptide peak could be clearly seen. Even then, this peak had only 39% of the area observed for columns constructed with the hybrid surface, suggesting that either more conditioning is required or that full recovery cannot be achieved using standard columns.
MS data were also obtained for the three-peptide mixture after separation using the conditioned columns, employing electrospray ionization and high-sensitivity quadrupole timeof-flight instrumentation. At first glance, we observed that the results from the total ion current chromatograms appeared to support the data acquired using UV detection, where there was little difference between the columns with respect to angiotensin I and enolase T37. However, upon further investigation, it was clear that higher quality MS data were obtained using the hybrid surface columns. Separations using standard columns yielded a relatively high ion signal for iron adducts, as exemplified in the mass spectra of enolase T37, shown in Figure 6. The mass spectrum obtained when using a standard column can be seen to be affected by iron ions that . Overlaid MS/MS-detected chromatograms for a mixture of ATP, ADP, AMP, and adenosine at a load of 50 pg of each analyte using a standard 1.8 μm HSS T3 column or a column containing the same packing material constructed using hybrid surface hardware (both 2.1 × 50 mm). The UHPLC system employed for this experiment had a flow path that was treated with the hybrid surface. ACN gradient separations were carried out using 10 mM NH 4 OAc (pH 6.8) as the aqueous mobile phase component.

Analytical Chemistry
pubs.acs.org/ac Article are leached from the stainless steel surface such that ironadducted peaks become an abundant feature ( Figure 6A).
Observations of iron adducts of peptides have previously been reported. 5 The levels of iron adducts that we observed for enolase T37 using a standard column were relatively low in the +2 and +3 charge states (see Supporting Information). However, for the +4 charge state, the abundance of the ironadducted peak was greater than that of the primary peak ( Figure 6A). In comparison, mass spectra obtained using columns constructed with the hybrid surface showed 80−90% reduced abundance of iron adducts ( Figure 6B). Interestingly, the charge state distribution for columns with the hybrid surface gave higher relative abundances for the lower charge states, suggesting that iron adduction affects ionization and forces analytes to occupy higher charge states. For example, the abundance of the +2 charge state relative to the +3 charge state of enolase T37 was approximately 50% higher when using a hybrid surface column versus a standard column (see Supporting Information). The presence of iron adducts makes the mass spectra more difficult to interpret due to the distortions in the relative abundances of protonated species and increased spectral crowding. 30 Evaluation for Oligonucleotide Separations. We also investigated the utility of UHPLC systems and columns constructed with the hybrid surface for the analysis of oligonucleotides. The analyte chosen for this study was trecovirsen, a 25-mer antisense oligonucleotide phosphorothioate that has been studied as a treatment for HIV-1. 31 The UHPLC system used for this experiment had components that were treated with the hybrid surface. An ion-pairing mobile phase containing 15 mM triethylamine (TEA), 400 mM HFIP, and methanol was employed. The same lot of the 1.7 μm oligonucleotide BEH C 18 stationary phase was packed in both standard columns and columns constructed using the hybrid surface. Three columns of each type were evaluated. The initial performance from the first four injections of 1.5 ng (193 fmol) of trecovirsen was evaluated before conditioning the columns with a high mass load (200 ng) of this oligonucleotide. A fifth injection was then made in order to determine whether the conditioning gave any improvements in the chromatographic performance.
Representative UV chromatograms resulting from the fourth and fifth injections are shown in Figure 7A,B. For the first injections on the standard columns, the average peak area of trecovirsen was 2−3 fold lower than that obtained using the hybrid surface columns ( Figure 7C). The relative standard deviations calculated from the results for the three standard columns were also greater by a factor of 8−17, as shown by the sizes of the ± one standard deviation error bars in Figure 7C. Moreover, the peak widths were decreased by 30% using the hybrid surface columns. After conditioning, the peak areas and peak widths improved by 30 and 6%, respectively, for columns packed in standard column hardware. Columns constructed with the hybrid surface showed reproducible performance both before and after column conditioning ( Figure 7C). Postconditioning changes in average peak area and peak width were less than 1% when using the hybrid surface columns. Even after conditioning the standard columns, the peak area of trecovirsen was still only 74% of the peak area observed using the hybrid surface columns.
Further differences in the chromatography were noticed in the resolution of co-eluting, lower abundant oligonucleotide species, as shown in Figure 7A,B. A later eluting peak, which was obscured by the main peak when using standard columns, could be partially resolved when using the hybrid surface Figure 5. UV chromatograms of the fourth injection (before conditioning) and fifth injection (after conditioning) of an equimolar mixture of doubly phosphorylated IR peptide (1), angiotensin I (2), and enolase T37 (3) obtained using a standard column (A) or a column constructed using hybrid surface hardware (B). Separations were performed with a 1.7 μm CSH C 18 stationary phase using an ACN gradient with 0.1% FA as the aqueous mobile phase and 20 pmol (25−50 ng) loads. The UHPLC system employed for this experiment had a flow path that was treated with the hybrid surface. Analytical Chemistry pubs.acs.org/ac Article Figure 7. UV chromatograms of the fourth injection (before conditioning) and fifth injection (after conditioning) of trecovirsen obtained using a standard column (A) or a column constructed using hybrid surface hardware (B) (both 2.1 × 50 mm). Separations were performed with a 1.7 μm Oligonucleotide BEH C 18 stationary phase using ACN gradients with TEA-HFIP-modified mobile phases and 1.5 ng mass loads. The UHPLC system employed for this experiment had a flow path that was treated with the hybrid surface. (C) Average peak areas of trecovirsen vs injection number obtained using standard columns or columns constructed using hybrid surface hardware (n = 3). The error bars show ± one standard deviation for three columns of each type.
Analytical Chemistry pubs.acs.org/ac Article columns. Through mass spectrometric analysis, performed with electrospray ionization and a tandem quadrupole mass spectrometer, this peak was identified as a +53 Da impurity of the main peak. This impurity likely arises from cyanoethylation of the oligonucleotide, a reaction which typically occurs on thymidines during the synthesis process. 32

■ CONCLUSIONS
The HST described here provides a means to improve UHPLC analyses of compounds that interact with metal surfaces. We have demonstrated that higher and more consistent peak areas and more symmetric peaks can be obtained using UHPLC systems and columns that incorporate this technology, especially for challenging analytes such as ADP, ATP, a doubly phosphorylated peptide, and a phosphorothioated oligonucleotide. Other phosphorylated analytes found to show improvement using this technology include phosphoglycans, sugar phosphates, and certain phospholipids. 26 In addition, significant benefits have been observed for analytes containing multiple carboxylate groups, such as acidic peptides and tricarboxylic acid cycle metabolites. 26,33 We have shown that hybrid surface UHPLC systems and columns give the greatest improvement over their standard counterparts when analyzing low-mass loads. This suggests that methods employing UHPLC/MS will benefit the most from this technology, particularly when trace level quantitative measurements are needed. The reduction in the occurrence of iron adducts, such as that noted in the mass spectra of enolase T37, is particularly important in studies using library matching. Work is in progress to further explore the range of applications that benefit from this technology.