Structure and magnetism of MnGe thin films grown with a nonmagnetic CrSi template
Extracted source text · arXiv 2510.13782
\documentclass[reprint,aps,twocolumn,prmaterials,superscriptaddress]{revtex4}
% MATH PACKAGES
\usepackage{gensymb}
\usepackage{amssymb}
\usepackage{bm, amsmath}
\usepackage{bm}% bold math
\usepackage{physics}
\usepackage[euler]{textgreek}
\usepackage{siunitx}
\usepackage[colorlinks=true,linkcolor=blue,urlcolor=blue,citecolor=blue]{hyperref}
\usepackage[version=4]{mhchem}
\usepackage{natbib}
\usepackage{siunitx}
% GRAPHICS PACKAGES
\usepackage{graphicx}% Include figure files
\usepackage[caption=false]{subfig}
\usepackage[outdir=./]{epstopdf}
\newcommand{\numero}{N\textsuperscript{\underline{o} }}
%\usepackage{times}
%TABLES
\usepackage{booktabs}
\usepackage{dcolumn}% Align table columns on decimal point
\usepackage{multirow}
%CITING
\usepackage{orcidlink}
%EDITING PACKAGES
\usepackage[normalem]{ulem} %strikeout command: \sout
\usepackage{verbatim}
\usepackage[draft]{todonotes}
\usepackage{float}
\usepackage{color}
%UNITS
\DeclareSIUnit{\QL}{QL}
\DeclareSIUnit{\Torr}{Torr}
\DeclareSIUnit{\bohr}{\mu_{\mathrm{B}}}
\DeclareSIUnit{\fu}{f.u.}
\begin{document}
\title{Structure and magnetism of MnGe thin films\\grown with a non-magnetic CrSi template}
\author{B.~D.~MacNeil\,\orcidlink{0000-0002-4731-3528}}
\affiliation{Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, B3H 3J5, Canada}
\author{J.~S.~R.~McCoombs\,\orcidlink{0009-0005-6955-6335}}
\affiliation{Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, B3H 3J5, Canada}
\author{D.~Kalliecharan\,\orcidlink{0000-0002-6429-279X}}
\affiliation{Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, B3H 3J5, Canada}
\author{J.~Myra\,\orcidlink{0009-0002-3087-2013}}
\affiliation{Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, B3H 3J5, Canada}
\author{M.~Pula\,\orcidlink{0000-0002-4567-5402}}
\affiliation{Department of Physics and Astronomy, McMaster University, Hamilton, Ontario, L8P 4N3, Canada}
\author{J.~F.~Britten\,\orcidlink{0000-0001-5527-0620}}
\affiliation{McMaster Analytical X-ray Diffraction Facility, McMaster University, Hamilton, Ontario, L8P 4N3, Canada}
\author{G.~B.~G.~Stenning}
\affiliation{ISIS Neutron and Muon Source, STFC Rutherford Appleton Laboratory, Didcot, OX11 0QX, United Kingdom}
\author{K.~Gupta\,\orcidlink{0000-0002-4826-6620}}
\affiliation{Catalan Institute of Nanoscience and Nanotechnology (ICN2), Campus UAB, Bellaterra, Barcelona, 08193, Spain}
\author{G.~M.~Luke\,\orcidlink{0000-0003-4762-1173}}
\affiliation{Department of Physics and Astronomy, McMaster University, Hamilton, Ontario, L8P 4N3, Canada}
\author{T.~L.~Monchesky\,\orcidlink{0000-0001-7401-7866}}\thanks{Contact author: tmonches@dal.ca}
\affiliation{Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, B3H 3J5, Canada}
%
\date{October 15, 2025}
\begin{abstract}
{We report on a novel method to grow B20 MnGe thin films which employs an ultrathin CrSi template layer on Si(111). This layer is expected to be non-magnetic, in contrast to MnSi and FeGe buffer layers that have been used previously, allowing an investigation of the intrinsic properties the MnGe in the ultrathin film limit without the influence of a neighboring magnetic layer. Single-phase MnGe(111) films were grown with thicknesses between 2 and \SI{40}{\nano\metre}, which exhibited low interfacial roughnesses on the order of \SI{0.6}{\nano\metre}. The films crystallized in a B20 structure with a small rhombohedral distortion. Magnetometry measurements in out-of-plane fields are consistent with a conical state. However, an unexpected remanent moment develops below \SI{35}{\kelvin}, concomitant with features in the field dependence of the transport data. This provides indirect evidence for the presence of a low-temperature phase which has been identified by others as either a triple-$Q$ topological spin-hedgehog lattice, or a multi-domain single-$Q$ conical state.}
\end{abstract}
%\pacs{}
\maketitle
\section{Introduction}
The B20 monosilicides and monogermanides comprise a rich phase diagram of magnetic states due to their broken inversion symmetry. The competition between symmetric exchange and the Dzyaloshinskii-Moriya interaction (DMI) results in helical magnetic order, with additional regions of the phase diagrams that host exotic magnetic textures~\cite{Dzyaloshinskii_1958, Moriya_1960, BakJensen_1980, Bogdanov_1989}. The prototypical B20 magnets MnSi and FeGe are known host a skyrmion state near their ordering temperature, where the magnetization twists into nanoscale cylindrical structures that are topologically protected~\cite{Muhlbauer_2009, Yu_2011}. The relative strength of symmetric exchange and DMI sets the wavelength $L_d$ of the helix and the lengthscale of the magnetic textures, which are typically much larger than the atomic spacing, \textit{e.g.}, approximately \SI{18}{\nano\metre} for MnSi and \SI{70}{\nano\metre} for FeGe.
On the other hand, MnGe exhibits the shortest helical wavelength among B20 compounds, which varies between approximately \SI{3}{\nano\metre} at low temperature, up to \SI{6}{\nano\metre} near the ordering temperature~\cite{Kanazawa_2011, Makarova_2012, Kanazawa_2012}. Electronic structure calculations show that the short helical wavelength is due to further-range exchange interaction~\cite{Mendive_2021}. While skyrmions have not been observed in MnGe, there are reports of a unique topological magnetic phase consisting of three-dimensional magnetic textures called spin-hedgehogs~\cite{Tanigaki_2015, Kanazawa_2016, Kanazawa_2017, Kanazawa_2020, Fujishiro_2021, Kitaori_2021, Hayashi_2021, Pomjakushin_2023}. Theoretical models describe it in terms of superposition of three orthogonal helical spin-density waves, referred to as a triple-$Q$ state, that can arise due to biquadratic exchange coupling \cite{Mendive_2021, Pomjakushin_2023}. The presence of this state was inferred from a large anomaly in the Hall resistivity attributed to a topological Hall effect, and Lorentz transmission electron microscopy (LTEM) measurements~\cite{Tanigaki_2015}. However, some studies claim that the triple-$Q$ phase is actually a topologically trivial a multi-domain single-$Q$ helical state with wavevectors aligned along each of the three $\langle 100 \rangle$ directions~\cite{Makarova_2012, Deutsch_2014, Altynbaev_2016, Yaouanc_2017, Altynbaev_2018, Martin_2019, Repicky_2021}. A multi-domain helical phase could also be responsible for the LTEM contrast observed in Ref.~\onlinecite{Tanigaki_2015}, but does not account for the large anomalies in the transport measurements.
The B20 phase of MnGe is metastable and preparing suitable samples is difficult. Polycrystalline MnGe was first synthesized at high pressure~\cite{Takizawa_1988, Kanazawa_2011} and the largest single crystals synthesized to date are around \SI{50}{\micro\metre} in size~\cite{Kanazawa_2020}. Epitaxial strain has also shown to be sufficient to stabilize MnGe. The first example utilized a \SI{1}{\nano\metre} seed layer of MnSi to grow MnGe on Si(111)~\cite{Engelke_2013}. Other groups have used a \SI{2}{\nano\metre} MnSi template~\cite{Kanazawa_2017, Kanazawa_2020, Fujishiro_2021, Hayashi_2021, Li_2025} or buffer layers of FeGe with thicknesses between 2 and \SI{10}{\nano\metre}~\cite{Ahmed_2017, Repicky_2021}. To date, all MnGe thin films have been grown using either MnSi or FeGe templates, which themselves are magnetic and can therefore influence the magnetic response of the MnGe. Previous studies have mainly focused on relatively thick MnGe films (\SI{80}{\nano\metre} to \SI{3}{\micro\metre}) and while Ref.~\onlinecite{Engelke_2013} studied films with thicknesses below \SI{13.5}{\nano\metre}, the magnetometry data was dominated by the moment of the MnSi template at low temperature.
\begin{figure*}[!htb]
\centering
\includegraphics[scale=1]{fig1.pdf}
\caption{(a) Unit cell of B20 MnGe. (b) Unit cell of rhombohedrally-distorted $R3$ MnGe shown for both in-plane tension ($\alpha>\ang{90}$) and compression ($\alpha<\ang{90}$). The arrows indicate the direction of the in-plane strain. (c)-(f) Illustration of the \si{\QL} stacking of the MnGe structure, for the four possible morphologies. The structures (c) and (e) comprise a left-handed crystal whereas (d) and (f) are right-handed. Crystal structures were visualized using the \textsc{vesta} software~\cite{Vesta}.}
\label{fig:MnGe_structure}
\end{figure*}
The magnetic structure of MnGe/MnSi/Si(111) films was probed by small-angle neutron scattering (SANS) in Ref.~\onlinecite{Kanazawa_2017}, which revealed single-$Q$ helimagnetic order above \SI{100}{\kelvin}. Below this temperature, the magnetic diffraction spot along $[111]$ split into two sets of three spots. For the \SI{1.8}{\micro\metre} thick film {($\approx$ bulk)}, these spots were along $\langle 100 \rangle$ directions. However, for the 160 and \SI{735}{\nano\metre} thick samples, the diffraction features were up to $\sim\ang{37}$ away from $\langle 100 \rangle$ towards the $[111]$. In these thinner samples, the alleged triple-$Q$ state was suppressed to a region below approximately \SI{40}{\kelvin} with the addition of a uniaxial anisotropy due to epitaxial strain from the substrate. Another report using spin-polarized scanning tunneling microscopy (SP-STM) on an \SI{80}{\nano\metre} MnGe/FeGe/Si(111) film found multi-domain single-$Q$ states, but no evidence for a triple-$Q$ hegdehog state. The intersection of these domains create topological objects, but they occur at a density lower than required to explain the features in the Hall effect data for MnGe.
In this work, we present the development of a B20 bilayer system designed to investigate the intrinsic magnetic properties of MnGe in the ultrathin film limit, where the helical wavevector is comparable to the film thickness. This system enables the interplay between finite size and large spin fluctuations, and the unique electronic structure of MnGe can be explored. We have characterized the structure of MnGe(111) grown on B20 CrSi template layers deposited Si(111) substrates, and investigated their properties with magnetometry and magnetotransport studies. We chose CrSi based on several studies that find no evidence of permanent magnetic order down to low temperature~\cite{Radovskii_1965, Mishra_2019}. Ref.~\onlinecite{Kousaka_2014} reports a weak moment of less than $0.04\mu_{\mathrm{B}}$ per Cr atom below \SI{20}{\kelvin} and Ref.~\onlinecite{Banik_2020} measured a moment of $0.01\mu_{\mathrm{B}}$ per Cr atom below \SI{2}{\kelvin} in an applied field of \SI{7}{\tesla}. Our magnetometry measurements indicate that the CrSi films are paramagnetic and are therefore not expected to have a large influence on the MnGe layer beyond potential changes to the electronic structure of the interfacial layer.
The rest of this paper is structured as follows: We first review the structure of B20 films on Si(111) substrates, and then describe the methods that we used to produce both a template of B20 CrSi, and B20 MnGe films using this template. We also discuss the effect of temperature and Mn composition in the stability of the B20 MnGe phase. We then outline the structural experiments we conducted to prove that we indeed grew epitaxial MnGe films, and to characterize the strain and distortion in these films due to the Si substrate. Finally, we investigate the magnetic properties of our films by presenting complementary magnetometry and transport data with a field applied parallel to MnGe$[111]$.
\section{Crystal Structure}
The crystal structure of B20 MnGe (space group \numero 198 $P2_13$) is detailed in Figure~\ref{fig:MnGe_structure}. The Mn and Ge atoms are located at the $4a$ Wyckoff sites and comprise a helix which winds around the $[111]$ directions according to the chirality of the crystal. The stacking of atoms in this direction can be described in terms of a quadruple layer (\si{\QL}), which has thickness corresponding to the (111) plane spacing of \SI{0.277}{\nano\metre}. A \si{\QL} consists of two dense and two sparse alternating layers of Mn and Ge atoms. The stacking unit along $[111]$ is comprised of twelve atomic layers, which can be described as ABC stacking of the \si{\QL}s. The stacking manner is determined by both the crystal chirality and its orientation. The MnSi/Si(111) system was found to nucleate domains with differing stacking sense, are related by either inversion, or \ang{180} rotation to give $[111]$ or $[\bar{1}\bar{1}\bar{1}]$ pairs~\cite{Trabel_2017, Morikawa_2020}. This yields four possible morphologies for MnGe/Si(111) when accounting for two crystal chiralities and two film orientations, which are denoted as \mbox{MnGe-L[111]}, \mbox{MnGe-L$[\bar{1}\bar{1}\bar{1}]$}, \mbox{MnGe-R[111]}, and \mbox{MnGe-R$[\bar{1}\bar{1}\bar{1}]$}. These stacking sequences are visualized in Fig.~\ref{fig:MnGe_structure}(c)-(f).
When a (111)-oriented B20 structure is subject to an isotropic in-plane strain, the cubic structure is distorted to a rhombohedral (space group \numero 143 $R3$) symmetry. In the distorted unit cell, the $4a$ Wyckoff sites of the Mn and Ge split into two inequivalent $1a$ and $3b$ Wyckoff sites. The rhomobhedral distorted unit cell is depicted in Fig.~\ref{fig:MnGe_structure}(b). Such a distortion was measured in thin film MnSi and \ce{Fe_{1-x}Co_xSi} grown on Si(111) with rhombohedral angles $\alpha$ of \ang{90.5} and \ang{92.5}, respectively~\cite{Karhu_2012, Porter_2012}. The distortion for FeGe was $\alpha=\ang{89.7}$ when grown on Si(111)~\cite{Turgut_2017} but $\alpha=\ang{94}$~\cite{Budhathoki_2020} when grown on Ge(111). The in-plane strain for these systems is expected to be tensile ($\alpha>\ang{90}$), whereas the expected lattice mismatch for MnGe(111)/Si(111) would be \mbox{$(2d_{1\bar{2}1}^{\mathrm{\,film}} - d_{1\bar{1}0}^{\,\mathrm{sub}})/2d_{1\bar{2}1}^{\mathrm{\,film}}+1.9\%$}, meaning pseudomorphic MnGe would be under compression.
\section{Film Growth}
MnGe thin films were prepared on high resistivity \mbox{($\rho>\SI{100}{\ohm\metre}$)} Si(111) substrates. The wafers were ultrasonically degreased for 15 minutes each in acetone and methanol, followed by a rinse and overflow in de-ionized nanopure water. The native oxide was etched for 10 minutes in a 1:2:10 solution of \ce{NH4OH}, \ce{H2O2} and nanopure water at \SI{70}{\degreeCelsius}, leaving the Si surface terminated by \ce{SiO2}. The wafer was again rinsed and overflowed, then dried with nitrogen gas and immediately loaded into the MBE system with base pressure less than \SI{5e-9}{\Pa}. The wafer and holder were degassed overnight at \SI{570}{\degreeCelsius} and the \ce{SiO2} was thermally desorbed at \SI{830}{\degreeCelsius} for 60 minutes. A \SI{20}{\nano\metre} Si buffer layer was deposited at \SI{640}{\degreeCelsius}, after which the sample was cooled to room temperature at a rate of less than \SI{1}{\degreeCelsius\per\second}. \emph{In-situ} reflection high energy electron diffraction (RHEED) revealed a strong $7 \times 7$ reconstruction [see Fig.~\ref{fig:rheed_afm}(a)-(b)], indicative of a pristine Si(111) surface which was also evidenced by \emph{ex-situ} atomic force microscopy (AFM) of the surface. AFM images of the substrate in Fig.~\ref{fig:rheed_afm}(c) shows step heights of approximately \SI{0.3}{\nano\metre}, which corresponds to the atomic spacing between Si(111) planes.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig2.pdf}
\caption{\textit{In-situ} RHEED and \textit{ex-situ} AFM images of (a)-(c) Si(111) substrate, (d)-(f) \SI{2}{\QL} \ce{CrSi} template (\SI{20}{\nano\metre} Si overlayer), (g)-(i) \SI{6.0}{\nano\metre} \ce{CrSi} film (\SI{20}{\nano\metre} Si overlayer), and (j)-(l) an \SI{7.2}{\nano\metre} \ce{MnGe} film (no Si overlayer) atop a \SI{2}{\QL} \ce{CrSi} template. The arrows in (e) indicate the fractional-order diffraction streaks visible in the $\surd 3 \times \surd 3$ pseudo-reconstruction of \SI{2}{\QL} CrSi with the beam along Si$[11\bar{2}]$. These are not present for the thicker CrSi film in (h). The RHEED patterns in each column share the same orientation relative to the substrate direction annotated in (a) and (b). All micrographs share the same scalebar.}
\label{fig:rheed_afm}
\end{figure}
We adapted the method of Higashi \emph{et al.}~\cite{Higashi_2009} to stabilize a \SI{2}{\QL} template of B20 CrSi. A trilayer with nominal thicknesses {$\SI{0.5}{\QL}$ Si/$\SI{1}{\QL}$ Cr/$\SI{0.5}{\QL}$ Si} was deposited at room temperature and then annealed under RHEED observation until a $\surd 3 \times \surd 3$ pseudo-reconstruction was achieved, typically after 60 minutes. The micrograph in Fig.~\ref{fig:rheed_afm}(f) reveals a smooth surface with with roughness
commensurate to that of the substrate.
%The step heights of the \SI{2}{\QL} CrSi template in Fig.~\ref{fig:rheed_afm}(f) were between \SI{0.51}{\nano\metre} and \SI{0.55}{\nano\metre}, in excellent agreement with the the CrSi(111) plane spacing.
After the room-temperature deposition of the Higashi trilayer, the Si $7\times 7$ reconstruction was replaced by diffuse RHEED spots representing the uncrystallized template. Upon annealing, sharp fractional-order spots [Fig.~\ref{fig:rheed_afm}(e)] appeared when beam was incident along Si$[1\bar{1}0]$, signaling the formation of B20 CrSi. Typical annealing temperatures were between 330 and \SI{410}{\degreeCelsius} for 30 to 75 minutes. When the CrSi trilayer was annealed above \SI{420}{\degreeCelsius}, the quality of the RHEED pattern dropped considerably and transmission spots indicated a rougher surface. This behavior is indicative of the formation of C40 \ce{CrSi2} (space group \numero 180 $P6_222$) which is reported to form in this temperature range with poor epitaxy until annealed above \SI{1100}{\degreeCelsius}~\cite{Wetzel_1987, Wetzel_1988, Mahan_1993}. Attempts at stabilizing a B20 CrSi template using a solid phase reaction between amorphous Cr and the Si $7\times 7$ surface either resulted in the formation of \ce{CrSi2} as evidenced by RHEED and later corroborated by \emph{ex-situ} X-ray diffraction (XRD), or no crystallization seen in RHEED (see the Supplemental Material for additional RHEED images of \ce{CrSi2}~\cite{supp}). Unless otherwise noted, all films were capped with an amorphous Si or Ge overlayer grown at room temperature, ranging from 4 to \SI{20}{\nano\metre} in thickness.
A two-stage annealing was also used to produce a thick film of B20 CrSi. Cr and Si were co-deposited onto the \SI{2}{\QL} template below \SI{100}{\degreeCelsius} and then annealed to \SI{350}{\degreeCelsius}
for approximately 60 minutes. The sharp diffraction patterns in Fig~\ref{fig:rheed_afm}(g)-(h) combined with micrograph in Fig~\ref{fig:rheed_afm}(i) are consistent with a flat surface with roughness below \SI{0.7}{\nano\metre} for a \SI{6}{\nano\metre} ($\approx\SI{22}{\QL}$) thick CrSi film. The XRD peak in Fig.~\ref{fig:xrd_MnGe}(b) shows well-resolved Laue fringes with roughness in agreement with AFM. Annealing temperatures exceeding this range would result in the formation of \ce{CrSi2} despite the \SI{2}{\QL} template being stable until approximately \SI{410}{\degreeCelsius}.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig3.pdf}
\caption{$\theta-2\theta$ XRD measurements of (a) the MnGe (111) peak for films grown atop \SI{2}{\QL} CrSi using the two-stage annealing procedure, and (b) the MnGe and CrSi (111) peak for a film grown atop a \SI{6}{\nano\metre} CrSi template using the same procedure. The inset in (b) shows phase purity of the B20 structure. Ticks in (c) denote the bulk peak positions for several crystal structures. Fits are performed using the model in Eq.~\ref{eq:xrd_peak}.}
\label{fig:xrd_MnGe}
\end{figure}
We found that co-deposition of Mn and Ge at lower temperatures ultimately led to smoother films. MnGe films were grown by co-deposition of Mn and Ge in a stoichiometric ratio onto \SI{2}{\QL} CrSi template layer at \SI{100}{\degreeCelsius}. At this point, the film produced an amorphous RHEED pattern. By annealing to \SI{250}{\degreeCelsius} for 60-90 minutes resulted in smooth crystalline surfaces as revealed by the diffraction patterns in Fig~\ref{fig:rheed_afm}(j)-(k). A similar method has also been employed with a seed layer of MnSi~\cite{Kanazawa_2017, Kanazawa_2020, Fujishiro_2021, Hayashi_2021, Li_2025}. The sharp streaks and low background are consistent with a flat multi-domain surface depicted by the micrograph in Fig~\ref{fig:rheed_afm}(l). Fig~\ref{fig:xrd_MnGe}(a) shows XRD acquired in a Siemens D500 diffractometer using Cu K$\alpha$ radiation for a variety of MnGe films grown with this two-stage annealing procedure. No impurity phases are detectable: only the MnGe (111) and (333) peaks are observed. The \SI{6}{\nano\metre} CrSi film also served as suitable template for MnGe with the same two-stage growth as the \SI{2}{\QL} template, depicted in Fig.~\ref{fig:xrd_MnGe}(b). The inset displays the full XRD pattern of this sample, which shows no evidence of any impurity phases such as \ce{CrSi2}.
We adapted a kinematical model of the Laue oscillations from Ref.~\onlinecite{Boulle_2006} to determine the film thickness and estimate the roughness of the interfaces. The scattering intensity is:
\begin{equation}
I(q) = \frac{|F|^2}{q^2}\qty[2-2\cos(qd)\exp\qty(-\frac{q^2({\sigma_1}^2 + {\sigma_2}^2)}{2})],
\label{eq:xrd_peak}
\end{equation}
where $q$ is the scattering vector, $F(q)$ is the form factor, $d$ is the film thickness, and $\sigma_1$ and $\sigma_2$ are the roughnesses at the upper and lower film interfaces, respectively. Accompanying fits to the MnGe(111) Laue fringes in Fig~\ref{fig:xrd_MnGe}(a) yielded interfacial roughnesses below \SI{0.8}{\nano\metre}. We also extracted the roughnesses from X-ray reflectometry (XRR) measurements that were fit using \textsc{genx}~\cite{GenX} and both methods were in agreement (see the Supplemental Material for XRR data and fitting~\cite{supp}). In Fig.~\ref{fig:params_vs_d}(d), we present interfacial roughness of the MnGe layer (whole film) as measured by XRR (AFM).
We also attempted to grow MnGe films atop the \SI{2}{\QL} CrSi template using the same recipe employed for MnSi, and the first synthesis of MnGe on Si(111)~\cite{Karhu_2011, Engelke_2013}. When Mn and Ge were co-deposited at \SI{250}{\degreeCelsius} and annealed for 60 minutes, XRD revealed a single peak at $2\theta = \ang{32.5}$ corresponding to MnGe(111) planes, although the absence of Laue oscillations flanking the main peak indicate that the MnGe film is relatively rough. Furthermore, a broad shoulder is visible near \ang{34} which indicates that some deposited Mn has reacted with the substrate to form MnSi. The same behavior was observed Ref.~\onlinecite{Engelke_2013} for a \SI{1}{\nano\metre} ($\approx\SI{3.8}{\QL}$) MnSi template. Increasing the temperature to \SI{325}{\degreeCelsius} led to more diffusion and a coexistence of epitaxial MnGe and textured MnSi phases which is replaced solely by the textured MnSi at \SI{400}{\degreeCelsius}. These findings are summarized in Fig~\ref{fig:xrd_Mn5Ge3}(a). When Mn and Ge were co-deposited on a \ce{CrSi2} template (created through either Si-rich deposition or Cr diffusion into the substrate above \SI{350}{\degreeCelsius}), no crystallization was observed in RHEED up to temperatures of approximately \SI{485}{\degreeCelsius}.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig4.pdf}
\caption{(a) $\theta-2\theta$ XRD measurements comparing the effect of temperature on the deposition of Mn and Ge atop \SI{2}{\QL} CrSi, leading to the formation of MnSi by diffusion above \SI{250}{\degreeCelsius}. (b) $\theta-2\theta$ XRD scans for Mn-rich films deposited on \SI{2}{\QL} CrSi. The as-grown film exhibits \ce{Mn5Ge3} impurities, which are removed when annealed to \SI{250}{\degreeCelsius} after annealing in excess Ge. This impurity phase remains upon heating above \SI{250}{\degreeCelsius}, where diffusion-formed MnSi is still observed in coexistence with MnGe and (smoother) \ce{Mn5Ge3}. The fit in (b) was performed using the model in Eq.~\ref{eq:xrd_peak}. Ticks in (c) denote the bulk peak positions for several crystal structures.}
\label{fig:xrd_Mn5Ge3}
\end{figure}
We also grew selected \mbox{Mn-Ge/\SI{2}{\QL} CrSi/Si(111)} samples with 10-15\% excess Mn on the \SI{2}{\QL} CrSi template layer. As seen in Fig.~\ref{fig:xrd_Mn5Ge3}(b), two-stage annealing to \SI{250}{\degreeCelsius} resulted in a mixture of B20 MnGe and D8\textsubscript{8} \ce{Mn5Ge3} (space group \numero 193 $P6_3/mcm$). Additional \textit{ex-situ} annealing for 60 minutes after deposition of a \SI{20}{\nano\metre} amorphous Ge overlayer is able to completely convert the Mn-rich impurity \ce{Mn5Ge3} into MnGe. Annealing these films to \SI{375}{\degreeCelsius} converted some of the excess Mn to B20 MnSi, presumably due to Mn diffusion through the template layer and into the substrate. Interestingly, the annealing improved the roughness of the \ce{Mn5Ge3} phase, as indicated by the emergence of Laue oscillations around the \ce{Mn5Ge3}(0002) peak. Mn concentrations as large as \ce{Mn_{5.0}Ge} deposited on the \SI{2}{\QL} CrSi template resulted in the formation of (0001)-oriented \ce{Mn5Ge3} rather than other Mn-rich polytypes. In the case of Mn-Ge films with excess Ge, we found that the \mbox{Mn-Ge/\SI{2}{\QL} CrSi/Si(111)} films are heterogeneous with a mixture of B20 MnGe(111) and Ge(111).
\section{Structure and Strain Characterization}
To gain further insight into the nanoscale film and interface structure, we performed cross-sectional scanning transmission electron microscopy (STEM) imaging. A thin lamella was prepared by focused ion beam and imaged in a Spectra 300 microscope. Figure~\ref{fig:tem_crosssection} shows two high angle annular dark field (HAADF) images of a \SI{22.9}{\nano\metre} thick MnGe film atop a \SI{2}{\QL} CrSi tempate and \SI{8}{\nano\metre} Si capping layer. The electron beam is parallel to Si$[11\bar{2}]$. The well-known \emph{staircase} pattern observed in Fig.~\ref{fig:tem_crosssection}(a) shows two domains with opposite stacking directions corresponding to opposite crystal chiralities. The symmetric selected area diffraction pattern (SADP) in Fig.~\ref{fig:tem_crosssection}(b) obtained by fast Fourier transform (FFT) over the entire film thickness is composed of two asymmetric diffraction patterns from each chiral domain. A wider view of the whole film is presented in Fig~\ref{fig:tem_crosssection}(c). The false color overlay is an elemental mapping of Mn, Ge, Cr and Si using energy-dispersive X-ray spectroscopy (EDS). The interfaces above and below the MnGe layer are sharp, and there is a distinct layer containing Cr at the film-substrate interface.
\begin{figure*}[ht!]
\centering
\includegraphics[scale=1]{fig5.pdf}
\caption{(a) HAADF-STEM image of a MnGe film depicting two chiral domains. (b) SADP calculated from the FFT of a lower magnification image of the domains in (a) encompassing the entire film thickness. (c) EDS elemental mapping of Mn, Ge, Cr, and Si. The substrate and capping layer are blue, and the MnGe film is shown in red. The orange layer at the Si-MnGe interface is the \SI{2}{\QL} CrSi template.}
\label{fig:tem_crosssection}
\end{figure*}
To further characterize the film structure, we performed a detailed investigation of reciprocal space using TEM and XRD. Samples for plan-view TEM were prepared by low-angle mechanical polishing and attached to TEM grids with silver epoxy as described in Ref.~\onlinecite{Robertson_2006}. Diffraction and imaging was done using the FEI Tecnai G2 F20 microscope. A sample SADP of the [111] zone-axis of a \SI{29.4}{\nano\metre} MnGe film is shown in Fig.~\ref{fig:tem_planview}(a). The fine structure of each spot is due to double-diffraction from the film and substrate~\cite{Karhu_2010}. We extracted the in-plane lattice spacings using the \textsc{fiji} software~\cite{Fiji}. The corresponding bright-field image in Fig.~\ref{fig:tem_planview}(c) appears uniform except for scratches due to the polishing, and free of impurity precipitates; similar to the AFM in Fig~\ref{fig:rheed_afm}.
To visualize the chiral domains seen in the HAADF-STEM images, we tilted the sample by approximately \ang{22.2} about Si$[2\bar{2}0]$ to reach the MnGe$[321]$ zone axis. Dark field imaging of the $(1\bar{1}0)$ and $(0\bar{2}1)$ spots in Fig.~\ref{fig:tem_planview}(b) reveal domains with sizes of approximately \SIrange{100}{800}{\nano\metre} like those in Fig.~\ref{fig:tem_planview}(d) across MnGe samples with thicknesses ranging from \SIrange{4}{40}{\nano\metre}. There is no preferential crystallographic orientation to the domain boundaries.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig6.pdf}
\caption{(a) Representative TEM SADP along the MnGe[111] zone axis for a \SI{29.4}{\nano\metre} MnGe film. The inset depicts the fine structure of the diffraction spots at higher magnification. (b) SADP of the MnGe[321] zone axis. (c) Bright field image along the MnGe[111] zone axis. (d) False color dark field image using the $(0\bar{2}1)$ spot (orange) and the $1\bar{1}1$ spot (blue) in (b) which reveals the chiral domains.}
\label{fig:tem_planview}
\end{figure}
We performed a more comprehensive investigation of reciprocal space by collecting X-ray reciprocal space maps (XRD-RSMs) of the films using a Bruker D8 Venture diffractometer with Cu K$\alpha$ radiation. We measured to a maximum $2\theta$ of approximately \ang{87} (\SI{56.3}{\per\nano\metre}) for one full $2\pi$ $\phi$-rotation and visualized the 3D dataset using the \textsc{max3d} software~\cite{max3d}. We used the diffraction spots from the Si substrate as an internal calibration and refined the film spots using the Bruker \textsc{apex} software and \textsc{xrayutilities} code~\cite{xrayutilities}. We refined two unit cells, beginning first with B20 MnGe. While this model roughly described the data, it did not accurately capture the peak locations across all of reciprocal space. The second model was the rhombohedral distortion of the B20 structure in Fig.~\ref{fig:MnGe_structure}(b) which gave significantly better agreement with the peak positions in the data, especially at larger $Q$. In Fig.~\ref{fig:xrd_rsm}(a) and (b), we present slices through the in-plane high-symmetry directions of the RSM for the same \SI{23}{\nano\metre} MnGe film in Fig.~\ref{fig:tem_crosssection}.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig7.pdf}
\caption{Slices through the XRD-RSM along the (a) Si$[11\bar{2}]$ and (b) Si$[1\bar{1}0]$ in-plane directions. Calculated peak positions for right- and left-handed rhombohedrally distorted MnGe are indicated by circles, along with squares for the substrate peaks. The size of the markers is proportional to the calculated intensity. (c) and (d) Pole figures centered around (111) for MnGe and Si \{111\} and \{400\} families. (e) and (f) Pole figures centered around (111) for MnGe \{210\} and \{220\} families.}
\label{fig:xrd_rsm}
\end{figure}
The calculated peak positions for rhombohedrally-distorted B20 MnGe are annotated in Fig.~\ref{fig:xrd_rsm} together with the Si (111) substrate peaks. Along the Si$[1\bar{1}0]$ direction, the RSM is symmetric and agrees with the SADP in Fig.~\ref{fig:tem_crosssection}(b), which is only explained by the superposition of two RSMs from both right- and left-handed MnGe. Figure~\ref{fig:xrd_rsm}(c)-(f) presents (111)-centered pole figures extracted from the XRD-RSMs. The \{111\} and \{400\} pole figures show that MnGe spots occur with a \ang{30} rotation relative to the substrate. The six MnGe spots correspond to three spots from two chiralities, like what was observed with SANS in Ref.~\onlinecite{Kanazawa_2017}.
The lattice constant $a$ and rhombohedral angle $\alpha$ of the distorted B20 films are given in Fig.~\ref{fig:params_vs_d}(a) and (b), calculated using both the TEM plane spacings and XRD-RSM refinement. All films studied exhibited $\alpha>\ang{90}$, similar to MnSi~\cite{Karhu_2012} and \ce{Fe_{1-x}Co_xSi}~\cite{Porter_2012} on Si(111). We also show the volume strain $\Delta V / V_{\textrm{bulk}}$ in Fig.~\ref{fig:params_vs_d}(c), which shows that the unit cell is larger than the bulk value above \SI{9}{\nano\metre} MnGe thickness, but there is a contraction of the unit cell for thinner films. We also comment that the TEM and XRD-RSM data generally agree, but the uncertainties in parameters obtained from the XRD-RSM are much lower.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig8.pdf}
\caption{(a) Rhombohedral lattice constant and (b) rhombohedral angle of MnGe vs film thickness, calculated from XRD RSM and TEM. The bulk lattice parameter of undistorted MnGe is noted ($\alpha=\ang{90}$). (c) Roughness of the MnGe layer vs film thickness, obtained by XRR and AFM. The XRR measures the roughness of the MnGe layer and AFM measures the roughness of the entire film. (d) Ordering temperature vs thickess, estimated from the maximum in the dc susceptibility at low fields ($\mu_0 H \leq \SI{100}{\milli\tesla}$). All MnGe films in this figure were grown with a \SI{2}{\QL} CrSi template, except the sample noted in (d) which used a \SI{6}{\nano\metre} template.}
\label{fig:params_vs_d}
\end{figure}
\section{Magnetic Characterization with out-of-plane field}
The magnetic moment of the films was measured with the MPMS-XL SQUID magnetometer with an applied field that was oriented parallel to the MnGe[111] direction. The diamagnetic effect of the Si substrate $\chi_{\rm{Si}}$ is typically obtained from the high-field portion of the $M$ vs $H$ loops, far above the saturation field. For the case of our films, this is not possible for two reasons. Firstly, the high-field susceptibility of bulk MnGe is small, but nonzero ($\chi_{\rm{HF}}\approx\SI{6}{\kilo\ampere\per\metre\per\tesla}$ at \SI{100}{\kelvin})~\cite{Kanazawa_2011}. Our previous work showed that MnSi films exhibit a similar $\chi_{\rm{HF}}$ to bulk~\cite{Meynell_2014} and we expect this to hold for MnGe. Secondly, the saturation field of our films is much larger than the maximum applied field of the magnetometer ($\pm$\SI{7}{\tesla}) at low temperatures. We therefore neglect the small temperature dependence of the low-temperature susceptibility and estimate $\chi_{\rm{Si}}$ from $M$ vs $H$ loops at temperatures where the film is saturated.
Representative $M$ vs $H$ loops for the \SI{22.9}{\nano\metre} thick MnGe film are given in Fig.~\ref{fig:squid_loops}(a) and (b). The form of the magnetization at high temperatures ($T>\SI{60}{\kelvin}$) is indicative of a conical phase and there is no hysteresis between increasing and decreasing fields. At low temperatures, ($T<\sim\SI{35}{\kelvin}$), there is a clear difference between increasing and decreasing magnetization branches. The largest difference \mbox{$\Delta M = M_{\mathrm{dec}} - M_{\mathrm{inc}} \approx \SI{16.5}{\kilo\ampere\per\metre}$} occurs at \SI{5}{\kelvin}. Figure~\ref{fig:squid_loops}(c) depicts $\Delta M$ and shows the temperatures and fields where it disappears. This hysteresis is not associated with a simple conical state, since the cone angle simply varies continuously with an field applied along $Q$. The remanent moment and $\Delta M$ are also too large to be explained by a small in-plane component of the applied field due to sample misalignment inside the magnetometer. The only feature in the $M$ vs $H$ loops is the second-order transition from the conical state into the saturated state. This transition occurs at a field $H_{\mathrm{c}_2}$ that is defined by the minimum in $\dv*[2]{M}{H}$.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig9.pdf}
\caption{(a) and (b) Corrected $M$ vs $H$ loops for the \SI{22.9}{\nano\metre} thick film with field applied along MnGe[111]. The only feature is the second-order transition between conical and saturated states. (c) Colormap showing the difference $\Delta M$ between decreasing and increasing magnetization branches. The squares indicate the field-temperature points where hysteresis disappears.}
\label{fig:squid_loops}
\end{figure}
The dc susceptibility $\chi=\dv*{M}{H}$ for constant applied fields is given in Fig~\ref{fig:M_rho_vs_T}(a). The peak in the susceptibility at \SI{10}{\milli\tesla} provides an estimate of the ordering temperature $T_{\mathrm{ord}}$ in the material. As the field increases the peak shifts to lower temperatures and corresponds to a transition from a conical state into a field-polarized ferromagnetic state. The thickness dependence of $T_{\mathrm{ord}}$ is shown in Fig.~\ref{fig:params_vs_d}(d). There is an enhancement of $T_{\mathrm{ord}}$ relative to bulk for all thicknesses presented here. The largest ordering temperature of \SI{206}{\kelvin} was observed for MnGe thicknesses of \SI{22.9}{\nano\metre} and \SI{29.4}{\nano\metre}. It has been reported that a \SI{160}{\nano\metre} MnGe film on a \SI{2}{\nano\metre} MnSi template layer also had an ordering temperature of \SI{206}{\kelvin}, and there was still an enhancement from bulk at \SI{1.8}{\micro\metre} MnGe thickness~\cite{Kanazawa_2017}.
\begin{figure}[!htb]
\centering
\includegraphics[scale=1]{fig10.pdf}
\caption{(a) Calculated dc susceptibility vs temperature for a \SI{9.1}{\nano\metre} MnGe film with $H\parallel[111]$. The ordering temperature is estimated by the position of the peak at low applied field. (b) Zero-field longitudinal and (c) transverse resistivity for the \SI{22.9}{\nano\metre} MnGe film. The secondary axes (lighter color) show the differential resistivity. The circles in (b) and (c) are values extracted from the resistivity-field measurements.}
\label{fig:M_rho_vs_T}
\end{figure}
We also conducted transport measurements using a DynaCool PPMS with maximum field of \SI{14}{\tesla}. The samples were cleaved into rectangles (nominally ${\SI{6.0}{\milli\metre}\times\SI{1.5}{\milli\metre}}$) and constant-current and transverse voltage leads were wire-bonded to the surface. The longitudinal $\rho_{xx}$ and transverse resistivity $\rho_{yx}$ was calculated using the field symmetrization and anti-symmetrization requirements, respectively. Figure~\ref{fig:M_rho_vs_T}(b) and (c) show the temperature dependence of $\rho_{xx}$ and $\rho_{yx}$ at zero-field. We applied the symmetrization and anti-symmetrization requirements to zero-field warmed data which was collected after field-cooling in $\pm\SI{14}{\tesla}$. Due to the complexity of symmetrizing temperature scans compared to field scans, we include values extracted from the $\rho$ vs $H$ measurements in the figure as a reference. There are features at low-temperature which are not observed in the magnetometry, including a reversal in the sign of the Hall resistivity at \SI{31}{\kelvin}. The $T_{\mathrm{ord}}$ estimated from features in the transport data is in agreement with the magnetometry.
The isothermal field dependence of $\rho_{xx}$ and $\rho_{yx}$ are shown in Fig.~\ref{fig:M_rho_vs_T}(a) and (c). The saturation field $H_{\mathrm{c}_2}$ is extracted from these data by examining features in the derivatives, as was done for the magnetization. There are additional features in the resistivity that are not observed in the magnetometry. There is a deviation from a field-linear behavior in $\rho_{yx}$ which is accompanied by a crossing of the decreasing and increasing field branches: a sign reversal in $\Delta\rho_{yx} = \rho_{yx}^{\mathrm{dec}}(H) - \rho_{yx}^{\mathrm{inc}}(H)$. One of the more conspicuous features is a maximum in $\dv*{\rho_{yx}}{H}$ below approximately \SI{40}{\kelvin} which also is also observed in $\dv*{\rho_{yx}}{T}$ at zero-field [cf. Fig.~\ref{fig:M_rho_vs_T}(c)]. We denote the fields of this feature as $H_{\mathrm{c}_1}$.
To address these anomalies and the topological nature of any low-temperature phases in MnGe films, we fit the transverse resistivity $\rho_{yx}$ to the a model from Ref.~\onlinecite{Meynell_2014_Hall} which accounts for ordinary and anomalous Hall effects. The Hall resistivity in this model is:
\begin{equation}
\rho_{yx} = R_0 \mu_0H + A M \rho_{xx} + B M {\rho_{xx}}^2 + \rho_{yx}^{\rm{other}}\,,
\label{eq:hall_resistivity}
\end{equation}
where $R_0 \mu_0H$ is the ordinary Hall effect, $A M \rho_{xx}$ accounts for extrinsic skew scattering, and $B M {\rho_{xx}}^2$ is a combined term describing extrinsic side-jump scattering and intrinsic contributions to the Hall resistivity. Due to the nonzero susceptibility above the saturation field and the inability to saturate the MnGe films at low temperatures, the Hall coefficient $R_0$ could not be determined prior to the fitting, as is commonly performed. Instead, we treated $R_0$ as a fitting parameter along with $A$ and $B$ (see the Supplemental Material for these fitting parameters~\cite{supp}). Alternatively, one study substituted the $R_0$ from a bulk polycrystal when fitting the Hall conductivity in MnGe films~\cite{Hayashi_2021}. The $\rho_{yx}^{\rm{other}}$ contribution is often attributed to a topological Hall effect $\rho_{yx}^{\rm{T}}$ due to the Berry phase that the electron acquires while following a magnetization texture with nontrivial topology such as a spin-hedgehog~\cite{Kanazawa_2011, Kanazawa_2012, Tanigaki_2015, Kanazawa_2016, Kanazawa_2017, Kanazawa_2020, Fujishiro_2021, Kitaori_2021, Hayashi_2021, Repicky_2021}. Additional contributions to the Hall effect in $\rho_{yx}^{\rm{other}}$ can arise from the helical structure~\cite{Lux_2020}, as found in MnSi films~\cite{Meynell_2014_Hall}. However, no evidence of such a contribution was found in MnGe films reported here.
\begin{figure*}[!htb]
\centering
\includegraphics[scale=1]{fig11.pdf}
\caption{(a) Transverse and (c) normalized longitudinal resistivity vs field for the \SI{22.9}{\nano\metre} thick MnGe film. (b) and (d) Phase diagram for this sample with $H\parallel[111]$. The circles are features derived from the resistivity and the squares are points where the hysteresis disappears in the $M$ vs $H$ data. The colormap in (b) is the difference between decreasing and increasing field branches of $\rho_{yx}$ and the colorlap in (d) is the unaccounted $\rho_{yx}^{\mathrm{other}}$ contribution to the resistivity after fitting Eq~\ref{eq:hall_resistivity} to the data. The closed (open) circle phase boundaries are $H_{\mathrm{c}_2}$ ($H_{\mathrm{c}_1}$).}
\label{fig:resistivity_all}
\end{figure*}
The colormaps Fig.~\ref{fig:resistivity_all}(b) and (c) depict $\Delta\rho_{yx}$ and $\rho_{yx}^{\rm{other}}$, along with the fields $H_{\mathrm{c}_1}$ and $H_{\mathrm{c}_2}$. The $\Delta M$ identified in Fig~\ref{fig:squid_loops}(c) are also shown, and align well with the $H_{\mathrm{c}_1}$ boundary. The phase diagram therefore appears to consist of three regions: a field polarized state above $H_{\mathrm{c}_2}$, a conical state between $H_{\mathrm{c}_1}$ and $H_{\mathrm{c}_2}$, and an additional low-temperature phase below $H_{\mathrm{c}_1}$. The significant hysteresis in the Hall measurements, signaled by nonzero $\Delta \rho_{yx}$ is largely confined to this region. The corresponding $\rho_{yx}^{\rm{other}}$ shown in Fig.~\ref{fig:resistivity_all}(d) extracted from Eq.~\ref{eq:hall_resistivity} does not show any significant contribution.
\section{Discussion}
Though it is likely not feasible to confirm the structure \SI{2}{\QL} of the CrSi template layers using XRD, the observed RHEED is sufficient to conclude that the amorphous {$\SI{0.5}{\QL}$ Si/$\SI{1}{\QL}$ Cr/$\SI{0.5}{\QL}$ Si} Higashi trilayer crystallizes upon annealing. While B20 CrSi and C40 \ce{CrSi2} may exhibit similar RHEED patterns, we note that the apparent drop in quality of the RHEED pattern with increasing temperature signals a structural change in the film. We suggest that this is due to diffusion of Si into the template which forms \ce{CrSi2}. As suggested in Refs.~\onlinecite{Wetzel_1987, Wetzel_1988, Mahan_1993}, CrSi/Si(111) is converted to \ce{CrSi2} accompanied by poor epitaxy which explains the appearance of transmission spots in the RHEED pattern. The fact that we were only able to grow MnGe or CrSi films atop the \SI{2}{\QL} CrSi is strong evidence that the template is indeed B20 CrSi, rather than another Cr-Si polymorph. We also wish to note that two recent studies reported the growth of (210)-textured monoclinic and cubic CrSi on Si(111) and claimed that the films exhibited ferromagnetic order between 3 and \SI{300}{\kelvin} including a moment of approximately 1$\mu_{\mathrm{B}}$ per Cr atom at \SI{300}{\kelvin}~\cite{Galkin_2025_1, Galkin_2025_2}. We do not observe this behavior (see the Supplemental Material for characterization of the \SI{6}{\nano\metre} CrSi sample~\cite{supp}).
The data presented in Figs.~\ref{fig:xrd_MnGe} and \ref{fig:xrd_rsm} show that B20 MnGe can be stabilized atop a template of B20 CrSi. To date, MnSi and FeGe have also been used. So long as the annealing temperature is kept below \SI{250}{\degreeCelsius}, diffusion of Mn into the Si substrate appears minimal based on XRD and TEM. This template is also capable of nucleating epitaxial B20 CrSi (while we found that a solid phase reaction could not reliably do so), and even D8$_{8}$ \ce{Mn5Ge3}. This invites the possibility of using varying thicknesses of a CrSi buffer layer to engineer strain in the overlaying film without impacting its magnetic properties. In particular, we note that the lattice mismatch for FeGe/Si(111) is $-0.31\%$ whereas it is $+1.28\%$ for FeGe/CrSi if the CrSi is fully relaxed.
Since the metastable MnGe structure does not appear to crystallize atop any template other than B20 MnSi, FeGe or CrSi, we hypothesize that during the post-deposition annealing step, crystallization initiates at the CrSi interface and proceeds upwards. Combining the RHEED, HAADF-STEM, and XRD-RSM data, we can conclude the following epitaxial relationship for our films:
\begin{align*}
&[1\bar{2}1]\ce{MnGe}(111) \parallel [1\bar{2}1]\ce{CrSi}(111) \parallel [1\bar{1}0]\ce{Si}(111).
\end{align*}
% This is redundant:
%&[10\bar{1}]\ce{MnGe}(111) \parallel [10\bar{1}]\ce{CrSi}(111) \parallel [11\bar{2}]\ce{Si}(111)\,.
% each material has two vectors that define the orientation unambiguously
%
The chirality of the MnGe domains is likely dictated by the underlying CrSi template. We observed evidence for two twin domains in our film, which are visualized directly in the opposite stacking directions of MnGe$[1\bar{2}1]$ planes in Fig.~\ref{fig:tem_crosssection}(a) and by dark-field imaging on the MnGe$[321]$ zone axis in Fig.~\ref{fig:tem_planview}(d). Our XRD-RSM measurements report similar findings to what Ref.~\cite{Trabel_2017} stated for MnSi, where it is clear that these twin domains are rotated by $\pm\ang{30}$ relative to the Si substrate and that the asymmetric intensities of the $\{210\}$ family in Fig.~\ref{fig:xrd_rsm}(e) indicate that there cannot be any rotational twinning (see the Supplemental Material for a calculation of pole figure intensities for the four morpholgies~\cite{supp}).
The calculated intensities of $\{210\}$ for the four stacking morphologies dictates that the following twins must occur together: \mbox{MnGe-R$[111]$} + \mbox{MnGe-L$[\bar{1}\bar{1}\bar{1}]$} or \mbox{MnGe-L$[111]$} + \mbox{MnGe-R$[\bar{1}\bar{1}\bar{1}]$}. If, for example, the film exhibited \mbox{MnGe-R$[111]$} + \mbox{MnGe-L$[111]$} domains, each spot in Fig.~\ref{fig:xrd_rsm}(e) would have equal intensity. In principle, another similar analysis of the $\{111\}$ reflection, for example, would allow us to distinguish which of these pairs is present in our films. However, the intensity asymmetry is too small to detect within the resolution of our measurements, nor can we elucidate the exact stacking manner within the resolution of our HAADF-STEM measurements. We note that Ref.~\onlinecite{Morikawa_2020} reported that \mbox{MnGe-L$[111]$} + \mbox{MnGe-R$[\bar{1}\bar{1}\bar{1}]$} was favored in MnSi/Si(111) using detailed reflectivity measurements.
The fact that we reported a rhombohedral angle \mbox{$\alpha>\ang{90}$} for all MnGe films is unexpected since the $+1.91\%$ epitaxial mismatch between MnGe and Si suggests compressive strain and therefore $\alpha<\ang{90}$. Strain data presented in Ref.~\onlinecite{Kanazawa_2017} for MnGe/\SI{2}{\nano\metre} MnSi is consistent with this unexpected behavior although no discussion about its origin was given. While both MnSi~\cite{Karhu_2012} and \ce{Fe_{1-x}Co_xSi}~\cite{Porter_2012} films on Si(111) also have a rhombohedral structure with $\alpha > 90$, this is expected based of their negative lattice mismatch. However, in the case of MnGe/Si(111) it is not clear why the positive lattice mismatch does not yield an in-plane compressive strain. In the case of MnSi/Si(111), Ref.~\onlinecite{Figueroa_2016} reports extended X-ray absorption fine structure (EXAFS) measurements that reveal larger relative changes to the positions of the Si atomic sites than to the unit cell itself. There are large differences in the coordination number between Si and MnSi. These differences could be driving changes in the bond angles near the MnSi/Si interface that are driving the rhombohedral distortion. Perhaps this is also true of the MnGe/CrSi/Si interface. We also note that the MnSi/SiC(0001) system is also expected to possess an in-plane compressive strain, but a tensile strain was found instead~\cite{Meynell_2016}. This was attributed to a dewetting of the film from the substrate and a thermal residual in-plane tensile strain after annealing to \SI{500}{\degreeCelsius}. Perhaps changes to MnSi bond angles near the MnSi/SiC interface are also influencing the strain in this system as well.
The magnetometry and transport data of our film is similar to the films presented in \cite{Kanazawa_2017, Fujishiro_2021, Hayashi_2021}, with a cone phase in most of the phase diagram below $T_{\mathrm{ord}}$ and subtle changes in the behavior below \SI{35}{\kelvin}. However, the additional contribution $\rho_{yx}^{\rm{other}}$ to the Hall effect in the \SI{22.9}{\nano\metre} MnGe film is only $\sim\SI{0.1}{\nano\ohm\metre}$ compared to $\sim\SI{1}{\nano\ohm\metre}$ in the thicker 80-\SI{300}{\nano\metre} films~\cite{Kanazawa_2017, Fujishiro_2021, Hayashi_2021}. It appears that the characteristic large negative contribution to $\rho_{yx}$ is thickness-dependent, since this signal is diminished for the {\SI{50}{\nano\metre} MnGe/\SI{5}{\nano\metre} FeGe/Si(111)} film in Ref.~\onlinecite{Repicky_2021} and largest for the bulk-like \SI{3}{\micro\metre} films~\cite{Kanazawa_2020, Kitaori_2021}. This reduction would be consistent with the interpretation given in Ref.~\onlinecite{Kanazawa_2017} that the reorientation of helical wavevectors toward the film normal leads to a more dilute distribution of emergent monopoles in the hedgehog lattice.
Nevertheless, we identify a low-temperature phase boundary defined by the onset of hysteresis in the magnetization, which coincides with the feature $H_{\mathrm{c}_1}$ in the Hall effect data. The location of the phase is consistent with the one defined by transport measurements in other films~\cite{Kanazawa_2017, Kanazawa_2020} where six-fold SANS diffraction patterns were observed. The field-dependent hysteresis has not been previously reported in this system. The origin of this hysteresis is not known, nor do we know what role surfaces and magnetocrystalline anisotropy are playing. A uniaxial compression of bulk polycrystalline MnGe causes a reorientation of the helical wavevector along the direction of the applied stress \cite{Deutsch_2014}. The $\alpha > 90$ rhombohedral distortion is therefore expected to drive the helical wavevector towards the $[111]$ direction, consistent with magnetometry at higher temperatures. As the temperature decreases, cubic anisotropy and anisotropic exchange would also be expected to increase in strength and contribute to a reorientation of the helical wavevector towards the $\langle 100 \rangle$ directions.
An additional possible contributing factor are finite size effects which can produce a canting of the helical propagation vector away from the $[111]$ direction to form the oblique spirals~\cite{Leonov_2020} observed in micromagnetic calculations of chiral magnetic films, but not yet measured experimentally.
\section{Summary}
We have demonstrated the growth of epitaxial MnGe(111) films on Si(111) using an ultrathin template of B20 CrSi. Contrary to those reported previously, the films investigated in this paper are free from the influence of a neighboring magnetic layer and have thicknesses comparable to the helical wavelength. A large portion of the magnetic phase diagram below the ordering temperature is a single-$Q$ conical state. We observed subtle hysteresis in the magnetization below a temperature of \SI{35}{\kelvin} that has not been reported and is coincident with features in the Hall resistivity. These signals indicate the presence of an additional phase, which is not observed in bulk. However, our magnetometry and transport measurements are not able to distinguish between multi-domain single-$Q$ states and triple-$Q$ spin-hedgehog states that are debated in the literature. A more direct probe such as neutron or resonant X-ray scattering is required to determine the magnetic structure of this low-temperature phase in the thin film limit.
\section*{ACKNOWLEDGMENTS}
This work was supported by the Natural Sciences and Engineering Research Council (Canada). We thank L. Kreplak at Dalhousie University for facilitating AFM measurements; V. Jarvis at MAX for assistance with X-ray RSM measurements; C. Lupien at Universit\'e de Sherbrooke for PPMS measurements; B. Ballesteros from ICN2 for assistance with TEM measurements; and M. D. Robertson at Acadia University for insightful discussions regarding TEM sample preparation and analysis.
We gratefully acknowledge the Science and Technology Facilities Council (STFC) for access to the ISIS Materials Characterisation Laboratory. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101007417 having benefitted from the access provided by ICN2 in Bellaterra (Barcelona) within the framework of the NFFA-Europe Pilot Transnational Access Activity, proposal ID697. We acknowledge the use of instrumentation as well as the technical advice provided by the Joint Electron Microscopy Center at ALBA (JEMCA) and funding from Grant IU16-014206 (METCAM-FIB) to ICN2 funded by the European Union through the European Regional Development Fund (ERDF), with the support of the Ministry of Research and Universities, Generalitat de Catalunya.
\section*{DATA AVAILABILITY}
The data that support the figures of this article are publicly available [to be added before publication]. Additional unshown and/or unprocessed data will be made available upon reasonable request.
\begin{thebibliography}{60}
\bibitem{Dzyaloshinskii_1958} I. Dzyaloshinskii, J. Phys. Chem. Solids \textbf{4}, 241 (1958).
\bibitem{Moriya_1960} T. Moriya, Phys. Rev. \textbf{120}, 91 (1960).
\bibitem{BakJensen_1980} P. Bak and M. H. Jensen, J. Phys. C \textbf{13}, L881 (1980).
\bibitem{Bogdanov_1989} A. N. Bogdanov and D. Yablonskii, Zh. Eksp. Teor. Fiz \textbf{95}, 178 (1989) [Sov. Phys. JETP \textbf{68}, 101 (1989)].
\bibitem{Muhlbauer_2009} S. M{\"u}hlbauer, B. Binz, F. Jonietz, C. Pfleiderer, A. Rosch, A. Neubauer, R. Georgii, and P. B{\"o}ni, Science \textbf{323}, 915 (2009).
\bibitem{Yu_2011} X. Yu, N. Kanazawa, Y. Onose, K. Kimoto, W. Zhang, S. Ishiwata, Y. Matsui, and Y. Tokura, Nat. Mater. \textbf{10}, 106 (2011).
\bibitem{Kanazawa_2011} N. Kanazawa, Y. Onose, T. Arima, D. Okuyama, K. Ohoyama, S. Wakimoto, K. Kakurai, S. Ishiwata, and Y. Tokura, Phys. Rev. Lett. \textbf{106}, 156603 (2011).
\bibitem{Makarova_2012} O. L. Makarova, A. V. Tsvyashchenko, G. Andre, F. Porcher, L. N. Fomicheva, N. Rey, and I. Mirebeau, Phys. Rev. B \textbf{85}, 205205 (2012).
\bibitem{Kanazawa_2012} N. Kanazawa, J.-H. Kim, D. S. Inosov, J. S. White, N. Egetenmeyer, J. L. Gavilano, S. Ishiwata, Y. Onose, T. Arima, B. Keimer and Y. Tokura, Physical Rev. B \textbf{86}, 134425 (2012).
\bibitem{Mendive_2021} E. Mendive-Tapia, M. dos Santos Dias, S. Grytsiuk, J. B. Staunton, S. Bl{\"u}gel, and S. Lounis, Phys. Rev. B \textbf{103}, 024410 (2021).
\bibitem{Tanigaki_2015} T. Tanigaki, K. Shibata, N. Kanazawa, X. Yu, Y. Onose, H. S. Park, D. Shindo, and Y. Tokura, Nano Lett. \textbf{15}, 5438 (2015).
\bibitem{Kanazawa_2016} N. Kanazawa, Y. Nii, X.-X. Zhang, A. Mishchenko, F. Kagawa, Y. Iwasa, and Y. Tokura, Nat. Commun. 7, 11622 (2016).
\bibitem{Kanazawa_2017} N. Kanazawa, J. White, H. R{\o}nnow, C. Dewhurst, D. Morikawa, K. Shibata, T. Arima, F. Kagawa, A. Tsukazaki, Y. Kozuka, M. Ichikawa, M. Kawasaki, and Y. Tokura, Phys. Rev. B \textbf{96}, 220414(R) (2017).
\bibitem{Kanazawa_2020} N. Kanazawa, A. Kitaori, J. White, V. Ukleev, H. R{\o}nnow, A. Tsukazaki, M. Ichikawa, M. Kawasaki, and Y. Tokura, Phys. Rev. Lett. \textbf{125}, 137202 (2020).
\bibitem{Fujishiro_2021} Y. Fujishiro, N. Kanazawa, R. Kurihara, H. Ishizuka, T. Hori, F. S. Yasin, X. Yu, A. Tsukazaki, M. Ichikawa, M. Kawasaki, N. Nagaosa, M. Tokunaga, and Y. Tokura, Nat. Commun. \textbf{12}, 317 (2021).
\bibitem{Kitaori_2021} A. Kitaori, N. Kanazawa, H. Ishizuka, T. Yokouchi, N. Nagaosa, and Y. Tokura, Phys. Rev. B \textbf{103}, L220410 (2021).
\bibitem{Hayashi_2021} Y. Hayashi, Y. Okamura, N. Kanazawa, T. Yu, T. Koretsune, R. Arita, A. Tsukazaki, M. Ichikawa, M. Kawasaki, Y. Tokura, and Y. Takahashi, Nat. Commun. \textbf{12}, 5974 (2021).
\bibitem{Pomjakushin_2023} V. Pomjakushin, I. Plokhikh, J. White, Y. Fujishiro, N. Kanazawa, Y. Tokura, and E. Pomjakushina, Phys. Rev. B \textbf{107}, 024410 (2023)
\bibitem{Deutsch_2014} M. Deutsch, P. Bonville, A. V. Tsvyashchenko, L. N. Fomicheva, F. Porcher, F. Damay, S. Petit, and I. Mirebeau, Phys. Rev. B \textbf{90}, 144401 (2014).
\bibitem{Altynbaev_2016} E. Altynbaev, S.-A. Siegfried, E. Moskvin, D. Menzel, C. Dewhurst, A. Heinemann, A. Feoktystov, L. Fomicheva, A. Tsvyashchenko, and S. Grigoriev, Phys. Rev. B \textbf{94}, 174403 (2016).
\bibitem{Yaouanc_2017} A. Yaouanc, P. Dalmas de R{\'e}otier, A. Maisuradze, and B. Roessli, Phys. Rev. B \textbf{95}, 174422 (2017).
\bibitem{Altynbaev_2018} E. Altynbaev, S.-A. Siegfried, P. Strau{\ss}, D. Menzel, A. Heinemann, L. Fomicheva, A. Tsvyashchenko, and S. Grigoriev, Phys. Rev. B \textbf{97}, 144411 (2018).
\bibitem{Martin_2019} N. Martin, I. Mirebeau, C. Franz, G. Chaboussant, L. N. Fomicheva, and A. V. Tsvyashchenko, Phys. Rev. B \textbf{99}, 100402 (2019).
\bibitem{Repicky_2021} J. Repicky, P.-K. Wu, T. Liu, J. P. Corbett, T. Zhu, S. Cheng, A. S. Ahmed, N. Takeuchi, J. Guerrero-Sanchez, M. Randeria, R. K. Kawakami, and J. A. Gupta, Science \textbf{374}, 1484 (2021).
\bibitem{Takizawa_1988} H. Takizawa, T. Sato, T. Endo, and M. Shimada, J. Solid State Chem. \textbf{73}, 40 (1988).
\bibitem{Engelke_2013} J. Engelke, D. Menzel, and V. Dyadkin, J. Phys. Condens. Matter \textbf{25}, 472201 (2013).
\bibitem{Li_2025} Z. Li, F. Meng, K. Zhao, X. Hua, Z. Huang, F. Quan, H. Ge, Z. Xiang, T. Wu, W. Meng, Y. Hou, Q. Lu, and X. Chen, Appl. Phys. Lett. \textbf{126}, 132403 (2025)
\bibitem{Ahmed_2017} A. S. Ahmed, B. D. Esser, J. Rowland, D. W. McComb, and R. K. Kawakami, J. Cryst. Growth \textbf{467}, 38 (2017).
\bibitem{Vesta} K. Momma and F. Izumi, J. Appl. Cryst. \textbf{44}, 1272 (2011).
\bibitem{Radovskii_1965} I. Radovskii, T. Shubina, P. Gel’d, and F. Sidorenko, Poroshk. Metall., \textbf{2} 26 (1965) [Sov. Powder Metall. Met. Ceram. \textbf{4} 115 (1965)].
\bibitem{Mishra_2019} A. K. Mishra , M. Krishnan, M. Gangrade , R. Venkatesh and V. Ganesan, AIP Conf. Proc. \textbf{2115}, 030436 (2019).
\bibitem{Kousaka_2014} Y. Kousaka, Y. Takahashi, N. Ikeda, H. Matsui, J. I. Kishine, Y. Togawa, M. Miyagawa, S. Nishihara, K. Inoue, and J. Akimitsu, JPS Conf. Proc. \textbf{3}, 014037 (2014).
\bibitem{Banik_2020} S. Banik, M. K. Chattopadhyay, S. Tripathi, R. Rawat, and S. N. Jha, Sci Rep \textbf{10}, 12030 (2020).
\bibitem{Trabel_2017} M. Trabel, N. V. Tarakina, C. Pohl, J. A. Constantino, C. Gould, K. Brunner, and L. W. Molenkamp, J. Appl. Phys. \textbf{121}, 245310 (2017).
\bibitem{Morikawa_2020} D. Morikawa, Y. Yamasaki, N. Kanazawa, T. Yokouchi, Y. Tokura, and T. Arima, Phys, Rev. Mater. \textbf{4} 014407 (2020).
\bibitem{Karhu_2012} E. A. Karhu, U. K. R{\"o}{\ss}ler, A. N. Bogdanov, S. Kahwaji, B. J. Kirby, H. Fritzsche, M. D. Robertson, C. F. Majkrzak, and T. L. Monchesky, Phys. Rev. B \textbf{85}, 094429 (2012)
\bibitem{Porter_2012} N. A. Porter, G. L. Creeth, and C. H. Marrows, Phys. Rev B \textbf{86}, 064423 (2012)
\bibitem{Turgut_2017} E. Turgut, A. Park, K. Nguyen, A. Moehle, D. A. Muller, and G. D. Fuchs, Phys. Rev. B \textbf{95}, 134416 (2017).
\bibitem{Budhathoki_2020} S. Budhathoki, A. Sapkota, K. M. Law, S. Ranjit, B. Nepal, B. D. Hoskins, A. S. Thind, A. Y. Borisevich, M. E. Jamer, T. J. Anderson, A. D. Koehler, K. D. Hobart, G. M. Stephen, D. Heiman, T. Mewes, R. Mishra, J. C. Gallagher, and A. J. Hauser, Phys. Rev. B \textbf{101}, 220405 (2020).
\bibitem{Higashi_2009} S. Higashi, P. Koc{\'a}n, and H. Tochihara, Phys. Rev. B \textbf{79}, 205312 (2009).
\bibitem{Wetzel_1987} P. Wetzel, C. Pirri, J. C. Peruchetti, D. Bolmont, and G. Gewinner, Phys. Rev. B \textbf{35}, 5880 (1987).
\bibitem{Wetzel_1988} P. Wetzel, C. Pirri, J.C. Peruchetti, D. Bolmont and G. Gewinner, Solid State Commun. \textbf{65}, 1217 (1988)
\bibitem{Mahan_1993} J. E. Mahan, R. G. Long, A. Vantomme, and M.-A. Nicolet, Mat. Res. Soc. Symp. Proc. \textbf{317}, 233 (1993).
\bibitem{supp} See Supplemental Material at [URL will be inserted by publisher]
for additional details about structural characterization and transport
measurements, calculation of XRD pole figure peak intensities, and the
XRD-RSM and mangetometry of the 6 nm CrSi sample.
\bibitem{Boulle_2006} A. Boulle, R. Guinebretière, and A. Dauger, Z. Kristallogr. Suppl. \textbf{23}, 123 (2006)
\bibitem{GenX} A. Glavic and M. Bj{\"o}rck, J. of Appl. Cryst, \textbf{55}, 1063 (2022).
\bibitem{Karhu_2011} E. A. Karhu, S. Kahwaji, M. D. Robertson, H. Fritzsche, B. J. Kirby, C. F. Majkrzak, and T. L. Monchesky, Phys. Rev. B \textbf{84}, 060404(R) (2011)
\bibitem{Robertson_2006} M. D. Robertson, T. Burns, and T. Morrison, Microsc. Soc. Can. Bull. \textbf{34}, 19 (2006).
\bibitem{Karhu_2010} E. Karhu, S. Kahwaji, T. L. Monchesky, C. Parsons, M. D. Robertson, and C. Maunders, Phys. Rev. B \textbf{82}, 184417 (2010).
\bibitem{Fiji} J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, et al., Nature methods 9, 676 (2012).
J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J.-Y. Tinevez, D. J. White, V. Hartenstein, K. Eliceiri, P. Tomancak, and A. Cardona Nat. Methods \textbf{9}, 676 (2012)
\bibitem{max3d} J. F. Britten and W. Guan, IUCr Comm. Cryst. Comp. Newsl. \textbf{8}, 96 (2007)
\bibitem{xrayutilities} D. Kriegner, E. Wintersberger, and J. Stangl, J. Appl. Cryst. \textbf{46}, 1162 (2013).
\bibitem{Meynell_2014} S. A. Meynell, M. N. Wilson, H. Fritzsche, A. N. Bogdanov, and T. L. Monchesky, Phys. Rev. B \textbf{90}, 014406 (2014).
\bibitem{Meynell_2014_Hall} S. A. Meynell, M. N. Wilson, J. C. Loudon, A. Spitzig, F. N. Rybakov, M. B. Johnson, and T. L. Monchesky, Phys. Rev. B \textbf{90}, 224419 (2014).
\bibitem{Lux_2020} F. R. Lux, F. Freimuth, S. Bl{\"u}gel, and Y. Mokrousov, Phys. Rev. Lett. \textbf{124}, 096602 (2020).
\bibitem{Galkin_2025_1} N. G. Galkin, K. N. Galkin, O. V. Kropachev, S. A. Dotsenko, D. L. Goroshko, D. B. Migas, A. B. Filonov, N. V. Skorodumova, A. V. Gerasimenko, and A. K. Gutakovskii, J. of Mater. Chem. C \textbf{13}, 2987 (2025).
\bibitem{Galkin_2025_2} N. G. Galkin, E. Y. Subbotin, K. N. Galkin, D. L. Goroshko, O. A. Goroshko, D. B. Migas, A. B. Filonov, I. A. Tkachenko, and A. Y. Samardak, J. of Mater. Chem. C \textbf{13}, 2875 (2025)
\bibitem{Figueroa_2016} A. I. Figueroa, S. L. Zhang, A. A. Baker, R. Chalasani, A. Kohn, S. C. Speller, D. Gianolio, C. Pfleiderer, G. van der Laan, and T. Hesjedal, Phys. Rev. B \textbf{94}, 174107 (2016).
\bibitem{Meynell_2016} S. A. Meynell, A. Spitzig, B. Edwards, M. D. Robertson, D. Kalliecharan, L. Kreplak, and T. L. Monchesky, Phys. Rev. B \textbf{94}, 184416 (2016).
\bibitem{Leonov_2020} A. O. Leonov, I. M. Tambovtcev, I. S. Lobanov, and V. M. Uzdin, Phys. Rev. B \textbf{102}, 174415 (2020).
\end{thebibliography}
\end{document}